The trillion-dollar club used to be a curiosity. When Apple first crossed the line in 2018, it stood alone. Today the threshold is crowded, and the crowd is overwhelmingly American. Eight of the ten most valuable companies on earth are now US-based, and in total ten American companies sit above the trillion-dollar line. The most recent arrival is not a software platform or a social network. It is Micron, a memory-chip maker that the market spent years treating as a commodity cyclical.

That detail is the tell. Three engines are running at once: artificial intelligence, cloud infrastructure, and now memory. The first two are familiar. The third is the one that confirms the build-out is real, because when the unglamorous layer of the stack starts producing trillion-dollar valuations, the demand underneath it is deep and durable.

The scoreboard

The numbers are stark. Nvidia sits above five trillion dollars. Apple and Alphabet are both above four and a half. Microsoft, Amazon, and Broadcom fill the next tier, with Tesla and Meta completing the American names in the global top ten. Only two outsiders break in: Taiwan’s TSMC, which makes the most advanced chips in the world, and Saudi Aramco, which still reflects the value of energy. Total global equity value now sits near 150 trillion dollars, up from roughly 90 trillion in 2020. The United States accounts for close to two-thirds of that figure while representing around a quarter of global output.

That gap between economic weight and market value is the premium investors are paying for American companies. It has widened, not closed. Concentration of this kind cuts both ways, and I will come back to the risk. First, the engine.

The engine: AI, cloud, and memory

The clearest evidence is what the largest buyers are spending. Microsoft, Amazon, Alphabet, and Meta plan to commit roughly 725 billion dollars of capital expenditure in 2026, up about 77 percent on the prior year, with most of it going to AI compute, data centres, and power. Analysts now expect the figure to clear one trillion dollars in 2027. Goldman Sachs models more than five trillion dollars of combined spending from these four firms across the second half of the decade.

Memory is where the story turns specific. Microsoft attributed a meaningful slice of its raised guidance to higher memory and component costs, and said it expects to stay capacity-constrained. Industry estimates put memory at roughly 30 percent of hyperscaler data-centre spending this year, several times its share two years ago. That demand lands on a small group of suppliers.

Micron’s June quarter shows what that looks like. Revenue reached 41.5 billion dollars, against 9.3 billion a year earlier. Gross margin printed at a company record near 85 percent. Free cash flow was 18.3 billion dollars. Guidance for the following quarter pointed to 50 billion dollars in revenue. Management framed it plainly, describing the strategic value of memory in the AI era and pointing to customer agreements that fix volumes and prices well into 2028. A business the market long valued as a price-taker is now signing the kind of long-dated contracts usually reserved for infrastructure.

This is the practical meaning of an AI build-out. It is not only chips and models. It is the entire physical stack, from lithography to logic to memory to power, and the value accrues to whoever owns the scarce layers.

The chasers, and the European gap

Look outside the United States and the picture thins quickly. The real chasers are Asian and Middle Eastern. TSMC sits above two trillion dollars. Aramco reflects energy. Samsung and SK Hynix, both Korean, are the other two pillars of the memory complex alongside Micron.

Europe is the conspicuous absence. The continent’s most valuable company is ASML, the Dutch firm whose lithography machines are a single point of dependence for the entire advanced-chip industry. It is indispensable, and it is worth roughly 620 billion dollars. That is a fraction of any American leader, and it sits below the line that ten US firms have crossed. Behind ASML, Europe’s next names, SAP, the Swiss pharma majors, the French luxury houses, are smaller still. A continent with comparable population and education has produced no member of the club.

The reasons are structural: fragmented capital markets, slower productivity growth, less risk capital, and a regulatory posture that taxes scale. The European gap is the clearest single piece of evidence that the trillion-dollar phenomenon is not a global technology story. It is a specifically American one.

Why this is endurance, not a sprint

The instinct is to call this a bubble. The more useful frame is endurance, in the exact sense the word carries in sport. Endurance is not a single explosive effort. It is the capacity to absorb stress and keep producing output, repeatedly, over time. By that test the American economy is performing.

US output grew around two percent in the first quarter of 2026, roughly four times the eurozone’s pace and ahead of every G7 economy except Canada. The World Bank attributed about two-thirds of its upward revision to global growth to American resilience. This happened during an oil shock tied to conflict with Iran and against persistent stagflation worries. The US absorbed the energy shock partly because it is now the world’s largest producer of oil and gas and a net exporter of both, which cushions exactly the kind of supply disruption that once caused recessions. The shock arrived, and the system kept running.

That is the endurance premium. Markets are not paying for a good quarter. They are paying for a structure that has repeatedly proven it can take a hit and keep compounding: deep capital markets, a culture of risk-taking, energy self-sufficiency, and a corporate sector that reinvests at scale.

Candour requires the other side. This resilience is concentrated and therefore fragile. PwC’s own economists call it teetering. Growth leans heavily on AI investment, supportive fiscal policy, and market confidence, and each of those has a limit. The top ten companies are a larger share of the market than at any point in modern history, so an AI disappointment would not stay contained. Memory remains cyclical, and a stock that has tripled can give back twenty percent quickly. US public debt is climbing toward levels that constrain future policy. None of this breaks the thesis. All of it sets the terms on which the thesis holds.

What it means for allocators

A few practical conclusions, offered as judgement rather than advice.

Own the compounders, but respect the cyclicals. The durable winners own scarce layers of the stack and reinvest cash at high returns. Memory rides the same cycle, but it is a different risk and should be sized as one.

Treat concentration as a position, not a backdrop. If you hold a broad US index, you already own this trade at maximum weight. The choice is whether to add to it or diversify away from it, and that choice should be deliberate.

Watch the engine, not the narrative. The number that matters is hyperscaler capital expenditure and the returns it produces. As long as that spending compounds and is justified by revenue, the cycle has fuel. The day the capex guidance stops rising is the day to pay close attention.

Do not confuse dominance with permanence. America owns this moment because of structure, not destiny. The same structure can erode, and Europe’s absence is a reminder of how fast an advantage compounds in one direction.

The trillion-dollar club is not a sign that markets have lost their minds. It is a measure of which economy can take a punch and keep building. For now, that economy is American. The discipline is to own the resilience without mistaking it for a guarantee.

Endurance Capital sits at the confluence of elite performance and capital allocation. If this was useful, subscribe and forward it to one allocator who will argue with it. New conversations with champions and the people building the frontier run through the season.

REFERENCES

1. Micron Technology, Inc. Record Results for the Third Quarter of Fiscal 2026. Press release and SEC Form 8-K, 24 June 2026. globenewswire.com; sec.gov.

2. CompaniesMarketCap. Largest Companies by Market Cap. Live ranking accessed 25 June 2026. companiesmarketcap.com.

3. Financial Times, compiled hyperscaler Q1 2026 capex guidance, via Tom’s Hardware and Yahoo Finance, April 2026.

4. Goldman Sachs Global Investment Research, hyperscaler capex modelling, via Yahoo Finance, 2026.

5. US Department of the Treasury. Economic Policy Statements to TBAC, 2026 Q2. home.treasury.gov, May 2026.

6. World Bank. Global Economic Prospects, January 2026. worldbank.org.

7. PwC. Annual Outlook 2026: Teetering Resilience. pwc.com, January 2026.

8. Morgan Stanley. Global Economic Outlook 2026: US Resilience to Lead Growth. morganstanley.com.

9. CaixaBank Research. The world economy in 2026; US 2026 outlook: resilience with frailties. caixabankresearch.com, November 2025.

10. US News & World Report. The Trillion-Dollar Club. money.usnews.com, June 2026.

All market capitalisation and price figures are point-in-time and move continuously. Figures reflect the dates and sources cited above.

Disclaimer

This article is provided for informational and educational purposes only. It reflects my analysis, interpretations, and opinions based on publicly available information, scientific literature, conference presentations, and the sources cited herein.

Nothing in this publication should be construed as investment advice, legal advice, tax advice, regulatory advice, medical advice, or a recommendation to buy, sell, or hold any security, company, fund, token, or other financial instrument.

Any investment decision should be based on independent research, due diligence, and consultation with qualified professional advisers where appropriate.

Scientific findings discussed in this article may evolve as additional research becomes available. References to longevity science, biological age, biomarkers, nutrition, exercise, artificial intelligence, or emerging healthcare technologies are intended solely for educational discussion and should not be interpreted as personal medical recommendations.

Past performance, whether financial, biological, or athletic, is not indicative of future results.

Editorial Note

This article was developed from my original notes, conference observations, source material, and independent research, and refined with the assistance of artificial intelligence as an editorial and research tool.

The ideas, analysis, conclusions, and final editorial decisions are my own. Any errors, omissions, interpretations, or judgments remain solely my responsibility.

I wanted to share with you something I learned during the 14th ISABS and Mayo Clinic Conference on Advances in the Application of Artificial Intelligence in Precision Medicine in Dubrovnik.

The conference brought together researchers working at the frontier of AI, genomics, aging, and precision medicine. There were presentations on machine learning models, multi-omics platforms, biomarkers, and the increasingly ambitious effort to quantify biological aging.

Yet one of the presentations that most captured my attention was delivered by Professor Claudio Franceschi.

Franceschi is widely recognized as one of the founders of the concept of inflammaging, the idea that chronic low-grade inflammation is a central feature of biological aging and a major contributor to age-related disease.

I arrived expecting to hear about artificial intelligence.

I left thinking about a much simpler idea.

The body is not one age.

The body is a mosaic of clocks

Most of us think of aging as a single process.

We celebrate birthdays. We count years. We use chronological age as a shorthand for biological reality.

But sitting in the audience in Dubrovnik, I was struck by how incomplete that view has become.

One of the slides presented research showing that organs and biological systems age at different rates. Using multi-omics data and machine learning, researchers are now constructing biological ages for individual systems, including the immune system, cardiovascular system, liver, kidney, skin, metabolism, and physical fitness.

A person may be 55 years old chronologically.

Yet their cardiovascular system may resemble that of a 40-year-old.

Their immune system may resemble that of a 70-year-old.

Their metabolic system may sit somewhere in between.

In other words:

The body is not a single clock. It is a portfolio of clocks.

This matters because organ-specific biological age may predict risk in the corresponding organ before disease becomes clinically visible.

The future of medicine may not ask:

“How old are you?”

It may ask:

“Which part of you is old?”

That is a fundamentally different question.

AI is making aging measurable

The conference was, after all, about artificial intelligence.

What fascinated me was not merely that AI models can estimate biological age.

It was that they are increasingly able to explain their reasoning.

Professor Franceschi discussed the use of explainable AI methods that move beyond prediction and toward interpretation.

Instead of simply generating a biological age score, these systems can identify which biomarkers contributed most to the result for a specific individual.

Medicine is slowly moving away from black-box averages and toward individualized biological explanations.

That distinction is important.

Prediction is useful.

Explanation is actionable.

CXCL9: The biomarker that kept appearing

One molecule repeatedly surfaced throughout the presentation: CXCL9.

Across multiple inflammatory aging models, CXCL9 appeared among the strongest contributors to biological aging estimates.

In one simplified immunological clock, AI reduced dozens of immune variables into a small group of key biomarkers. CXCL9 ranked first.

In inflammatory aging indices, it again emerged as one of the strongest positive contributors.

Explainable AI analyses consistently showed CXCL9 exerting a meaningful influence on whether an individual’s predicted biological age moved higher or lower.

This does not mean CXCL9 causes aging.

The science does not support such a claim.

But it does suggest that CXCL9 may be one of the clearest molecular fingerprints of inflammaging currently emerging from the data.

As investors and operators, we often look for recurring signals.

When a variable appears across datasets, models, and methodologies, it deserves attention.

CXCL9 appears to be one of those signals.

The most sophisticated biomarkers led to surprisingly simple interventions

What surprised me most was not the sophistication of the AI models. It was the simplicity of many of the proposed interventions.

One slide mapped major drivers of inflammaging to potential actions:

• Sedentary lifestyle → regular exercise

• Western diet → Mediterranean-style nutrition

• Sleep disturbance → improved sleep hygiene

• Smoking and alcohol → reduction or cessation

• Social and environmental factors → broader health and welfare interventions

The biomarkers were sophisticated. The first interventions were not. This was perhaps the most important takeaway from the session. The modern longevity industry often gravitates toward complexity. New molecules. New diagnostics. New therapies. Yet some of the strongest evidence discussed in Dubrovnik continues to point toward fundamentals.

Exercise. Sleep. Nutrition. Social connection.

These are often dismissed as lifestyle variables.

The emerging science increasingly suggests they are biological variables.

The mediterranean diet Is more than a diet

Several studies presented during the conference focused on the Mediterranean diet and aging.

Among them was the NU-AGE project, which found that a one-year Mediterranean diet intervention was associated with reduced inflammaging and measurable epigenetic changes in older adults.

Another study linked Mediterranean diet adherence to beneficial shifts in the gut microbiome, reductions in frailty, and improvements in cognitive outcomes.

The details will continue to evolve.

The broader lesson is already clear.

A relatively low-cost behavioral intervention appears capable of influencing multiple biological systems simultaneously.

Inflammation.

Microbiome composition.

Epigenetic aging.

Frailty.

Cognition.

From an investment perspective, this is a useful reminder. Biology is a network. The most powerful interventions may not act on a single pathway. They may influence several systems at once.

Social connection is biology

One of the most unexpected parts of the presentation involved social relationships.

Research discussed during the session linked loneliness and social isolation with gut microbiome disruption, systemic inflammation, frailty, and cognitive decline.

Conversely, social engagement appeared associated with healthier microbiome patterns, cognitive stimulation, and healthier aging trajectories.

We often think about social connection as psychology.

The emerging evidence suggests it may also be biology. That distinction matters. Because it changes how we think about prevention.

What this means for investors

Every major technological cycle creates new measurement systems. The internet created analytics. Mobile created engagement metrics. Cloud computing created observability. Longevity may create biological observability. The largest opportunities may not belong to companies promising to cure aging. They may belong to companies that can measure, explain, and modify the biological systems that drive aging before disease appears.

Organ-specific aging diagnostics.

Inflammatory biomarker platforms.

Biological age infrastructure.

AI-powered preventive medicine.

Personalized health monitoring.

The winners may not extend life directly. They may help make aging measurable.

An Endurance Capital reflection

As I left the conference hall in Dubrovnik, I kept returning to a simple analogy from endurance sport.

An endurance athlete does not wait until race day to understand performance.

Training load. Recovery. Nutrition. Sleep. Heart-rate variability. Resilience.

The race result is merely the final output.

Medicine has traditionally focused on the equivalent of race-day outcomes. Heart attacks. Diabetes. Frailty. Dementia.

The biological clocks described by Professor Franceschi point toward a different future.

Disease is the outcome.

Biological age is the process.

And the future of medicine may increasingly be about measuring that process before the outcome arrives.

That is an important scientific shift.

It may also become an important capital allocation opportunity.

Key Takeaways

1. The body is not one age. Different organs and systems age at different rates.

2. Organ-specific biological age may become a more useful metric than chronological age.

3. Explainable AI is helping researchers understand why biological age predictions are made.

4. CXCL9 is emerging as one of the strongest recurring signals associated with inflammaging.

5. Exercise, nutrition, sleep, and social connection remain among the most evidence-supported interventions.

6. Biological age measurement may become a major category in preventive medicine.

7. The most attractive investment opportunities may emerge from companies that measure and explain aging rather than merely treating disease.

References

Nie, C. et al.
Multi-omics-based aging clocks for organs and biological systems.
Cell Reports. 2022.

Franceschi, C. et al.
Inflammaging and age-related disease.
Nature Aging and related geroscience publications.

Sayed, N. et al.
An inflammatory aging clock (iAge) based on deep learning tracks multimorbidity, immunosenescence, frailty and cardiovascular aging.
Nature Aging. 2021.

Gensous, N. et al.
One-year Mediterranean diet promotes epigenetic rejuvenation with country- and sex-specific effects: a pilot study from the NU-AGE project.
Geroscience. 2020.

Ghosh, T.S. et al.
Mediterranean diet intervention alters the gut microbiome in older people, reducing frailty and improving health status.
Gut. 2020.

Biagi, E. et al.
Gut Microbiota and Extreme Longevity.
Current Biology. 2016.

Ticinesi, A. et al.
Social networks, microbiome health, and healthy aging.
Ageing Research Reviews. 2025.

Disclaimer

This article is provided for informational and educational purposes only. It reflects my analysis, interpretations, and opinions based on publicly available information, scientific literature, conference presentations, and the sources cited herein.

Nothing in this publication should be construed as investment advice, legal advice, tax advice, regulatory advice, medical advice, or a recommendation to buy, sell, or hold any security, company, fund, token, or other financial instrument.

Any investment decision should be based on independent research, due diligence, and consultation with qualified professional advisers where appropriate.

Scientific findings discussed in this article may evolve as additional research becomes available. References to longevity science, biological age, biomarkers, nutrition, exercise, artificial intelligence, or emerging healthcare technologies are intended solely for educational discussion and should not be interpreted as personal medical recommendations.

Past performance, whether financial, biological, or athletic, is not indicative of future results.

Editorial Note

This article was developed from my original notes, conference observations, source material, and independent research, and refined with the assistance of artificial intelligence as an editorial and research tool.

The ideas, analysis, conclusions, and final editorial decisions are my own. Any errors, omissions, interpretations, or judgments remain solely my responsibility.

Or listen wherever you get your podcasts.

There is a version of high performance that everyone sees.

The victory.

The record.

The company milestone.

The moment when years of work become publicly visible.

But the result is not the system.

It is only evidence that the system worked.

My guest on Endurance Capital is Chrissie Wellington OBE, four-time IRONMAN World Champion and one of the most accomplished athletes endurance sport has produced.

When I asked what enabled her success, she did not begin with talent, physiology or trophies.

She began with something more useful:

The system is the success.

The medal is only its visible manifestation.

Sustained excellence is not one exceptional performance. It is the ability to prepare, perform, learn and return.

For Chrissie, that system begins with purpose.

Before the training plan, there must be a reason to train. Before the goal, there must be an understanding of what matters.

Her goals changed throughout her career. She moved from recreational triathlete to professional athlete, and from wanting to finish near the front to becoming World Champion.

The goals evolved.

The underlying purpose remained.

This matters in company-building too.

Products change. Markets move. Strategies need to be rewritten.

A strong system does not defend the original plan at all costs. It protects the mission while allowing the method to adapt.

Chrissie calls this consistency with agility.

You continue doing the work, but remain responsive to what the body, the data and the environment are telling you.

You act.

You review.

You recalibrate.

Then you act again.

Consistency is not repeating the same plan indefinitely. It is remaining committed while adapting to reality.

Confidence develops in much the same way.

We often assume that champions and founders act because they possess unusual certainty.

But Chrissie did not arrive at her first World Championship convinced she would win. Her goal was a top-ten finish.

She won because she focused on the process and executed the system she had built.

Confidence came through preparation, repetition and experience.

Failure contributed too.

Setbacks can show us that we are more adaptable than we believed. They create evidence that uncertainty can be survived.

This produces a more durable form of confidence.

Not the belief that everything will go to plan.

The belief that you can respond when it does not.

Confidence is built in the trenches of experience. Action creates evidence; evidence changes belief.

But high performance is not simply about pushing harder.

One of the most important lessons from Chrissie’s career is that not every limit should be broken.

Some limits are psychological. They are built from fear, habit or an incomplete understanding of what we can do.

These deserve to be tested.

Other limits are structural.

Sleep.

Health.

Energy.

Time.

Cognitive capacity.

These protect the system.

Ignoring them may produce a short period of exceptional output, but it eventually weakens the ability to perform at all.

Chrissie experienced both.

In 2010, she withdrew from the World Championship because she was unwell. The strongest decision was to respect the signal.

A year later, she reached the same race after a serious bicycle accident disrupted her preparation and left her injured.

She focused on execution and won her fourth world title.

The lessons are not contradictory.

In one case, courage meant stopping.

In the other, it meant beginning despite uncertainty.

The skill was interpreting the signals accurately.

Some limits are assumptions to challenge. Others are foundations to protect. Wisdom lies in knowing the difference.

This is why health sits at the centre of Chrissie’s performance system. She describes life as a pie. Work, family, relationships and interests are the slices. Health is the crust. Without it, the rest loses its structure.

The endurance lesson is not that recovery replaces hard work. It is that stress and recovery form one productive cycle. Training creates the stimulus. Recovery creates the adaptation.

The same applies to companies and capital. Growth without organisational recovery creates fragility. Investment without sufficient reserves removes optionality. Intensity without reflection reduces decision quality. Sometimes the system needs space to learn.

Chrissie also spoke about environment. Individual agency matters, but nobody performs alone. Coaches, family, training conditions and access to the right resources all helped make her career possible. The right environment does not remove responsibility. It makes ambition sustainable.

Perhaps the deepest part of the conversation concerned identity. Winning matters. But when a person’s identity becomes completely attached to one result, company, title or achievement, the system becomes unstable.

Chrissie remains proud of her victories, but she is no longer defined by them.

A founder must remain larger than the company.

An investor must remain capable of changing their mind.

An athlete must eventually discover meaning beyond competition.

Achievement should become part of the person.

It should not contain the whole person.

Near the end of our conversation, I described sustained high performance as the goal.

Chrissie corrected me.

Sustained high performance is not the goal. It is the process.

That may be the central lesson.

The objective is not to remain permanently at peak intensity.

Nothing living can do that.

The objective is to build a system that can prepare, perform, recover, learn and return.

Again and again.

With consistency.

With agility.

And without destroying the foundations that made the performance possible.

————————————————————

Key lessons for founders, investors and high-performers

The system is the success.

The result is only one visible expression of it.

Confidence follows action.

Preparation and experience create more durable confidence than manufactured certainty.

Practise consistency with agility.

Protect the mission, but remain willing to change the plan.

Differentiate between limits.

Challenge psychological assumptions while protecting health, energy and decision quality.

Remain larger than the outcome.

Achievement should shape identity, not imprison it.

————————————————————

Disclaimer

This article is educational and experience-based. It is not medical advice, training advice, investment advice or a substitute for professional guidance.

Any discussion of endurance training, performance, recovery, leadership or capital allocation is based on personal experience and general reflection, not individual prescription.

Always consult qualified professionals before making material changes to your health, training, recovery or investment decisions.

Sponsor relationships are disclosed where relevant.

————————————————————

Editorial note

Endurance Capital explores the intersection of endurance sport, founders, longevity and capital allocation.

This episode features Chrissie Wellington, four-time IRONMAN World Champion and one of the most accomplished endurance athletes of her generation.

Hosted by Ignacio Garcia, founder of Trampoline Venture Partners, Endurance Capital translates lessons from elite performance into practical frameworks for people building, leading and backing durable things.

THE G7 AI MOMENT

Why the quiet presence of Amodei, Altman and Hassabis at Évian matters for markets, safety and the next productivity cycle

The most important signal from the G7 meeting in Évian-les-Bains was not a speech.

It was the presence of three people in the room.

Dario Amodei. Sam Altman. Demis Hassabis.

The CEOs and scientific leaders behind Anthropic, OpenAI and Google DeepMind were not there as keynote entertainment. They were sitting with the leaders of the world’s most powerful democratic economies at a moment when artificial intelligence is moving from a technology story into a strategic infrastructure story.

AI is no longer just a software cycle. It is becoming strategic infrastructure.

That is a quiet but important upgrade.

It means AI is now being treated less like another software cycle and more like energy, semiconductors, defense, cyber security and financial stability: something too important to leave entirely to market forces, and too fast-moving to regulate from a distance.

The Financial Times framed the moment clearly. Amodei urged G7 leaders to “resist the temptation to splinter,” shortly after the US restricted foreign access to Anthropic’s latest models on national-security grounds. Altman and Hassabis also supported the case for closer cooperation between democratic countries on frontier AI development, evaluation and safety.

That is the line that matters:

resist the temptation to splinter.

Because in AI, fragmentation is not just inefficient.

It is dangerous.

For the last two years, the public conversation around AI has mostly been about capability.

Which model is smarter?

Which company is ahead?

Which country has the chips?

Which applications will transform work first?

Those questions still matter. But the G7 meeting points to a different question now becoming central:

Can the democratic world coordinate fast enough to make AI adoption safe, trusted and durable?

That is not a soft policy question. It is a market question.

If AI develops through mistrust, export shocks, unilateral restrictions, incompatible national rules and cyber escalation, adoption slows. Boards become cautious. Banks and insurers delay deployment. Governments become defensive. Companies become unsure whether the tools they build on today will still be available tomorrow.

Uncertainty is a tax on adoption. In AI, mistrust becomes a drag on growth.

That kind of uncertainty becomes a tax on growth.

The recent Anthropic dispute made that visible. Reuters reported that G7 leaders discussed closer AI coordination and a possible “trusted partners” approach after the US restricted foreign access to Anthropic’s Fable 5 and Mythos 5 models. Macron warned that companies would stop buying American AI if Washington could switch it off at will, and that without transatlantic cooperation the world would simply fracture. Ursula von der Leyen said it was in the mutual interest of the US and Europe for Europe to access the best AI models safely.

The Wall Street Journal also treated the issue as more than a company dispute, reporting that President Trump said negotiations with Anthropic were “going fine” while other countries worried about losing access to top US AI systems.

Le Monde placed the same moment in a broader security frame, connecting the G7 discussions in Évian with parallel UN consultations in Geneva on military AI. Its reporting highlighted the deeper concern: AI is becoming relevant not only to productivity and commerce, but to war, targeting, surveillance, cyber defense and international humanitarian law.

So the real story is not simply that three AI leaders attended a G7 lunch.

The real story is that AI has crossed a threshold.

It is now part of the operating system of state power.

There is a temptation to read this only as regulation.

That would be too narrow.

The better reading is that democratic governments are trying to build a safety net under the next productivity cycle.

Safety is not the opposite of growth. At this stage, safety is what allows growth to scale.

That safety net has several parts.

First, access.

European and allied economies do not want to build their future productivity stack on tools that can be switched off overnight. That is not an irrational fear. If companies, banks, health systems or infrastructure providers become dependent on frontier AI, continuity of access becomes strategically important.

Second, security.

The same models that improve cyber defense can also be misused. The same reasoning capabilities that help researchers may also lower the cost of harmful activity. This is why export controls, model evaluations and trusted-access regimes are now part of the conversation.

Third, standards.

If each country develops a different framework for frontier models, the result will not be safety. It will be complexity. The companies that can afford compliance teams will survive. Smaller companies will slow down. Cross-border deployment will become harder. Innovation will become more regional, more political and less efficient.

Fourth, trust.

This is the part markets often underestimate. AI adoption depends on confidence. Customers need to believe the tools are reliable. Governments need to believe the companies are responsible. Allies need to believe they will not be cut off. Investors need to believe the regulatory path is navigable.

Without trust, capability does not translate cleanly into productivity.

It becomes trapped in pilots, committees and restricted use cases.

This is why the presence of Amodei, Altman and Hassabis matters.

Each one represents a different part of the frontier AI conversation.

Amodei represents the safety-first view: the argument that model capability is advancing quickly enough that governments must take catastrophic risk, cyber risk and misuse seriously.

Altman represents scale and deployment: the belief that AI will become a broad economic platform, and that democratic countries need institutions capable of governing it without choking it.

Hassabis represents deep scientific capability: the long arc from research into systems that may reshape biology, medicine, materials, energy, education and defense.

Together, they signal that the conversation has moved beyond “how do we regulate tech companies?” to something more structural:

How do democratic societies govern a capability that is becoming part of their economic and security base?

That is a much harder question.

It is also the right one.

The next AI winners may not only be the fastest builders. They may be the most trusted builders.

For financial markets, this should be read as positive.

Not because the G7 solved AI governance. It did not.

Not because every restriction is good. Some may be badly designed.

But because serious coordination is better than fragmentation.

AI is likely to become a major driver of efficiency across software, finance, health, logistics, research, industrial operations and government services. But that efficiency will only compound if adoption is broad, trusted and resilient.

An AI cold war between allies would slow the cycle.

A credible cooperation framework would support it.

This is why safety is not anti-growth.

At this stage, safety may be what allows growth to scale.

The mistake is to imagine a trade-off between speed and cooperation. In systems this powerful, cooperation is part of speed. It reduces uncertainty. It lowers adoption friction. It gives enterprises and governments a clearer basis for investment.

In endurance terms, this is the difference between reckless pace and sustainable pace.

The fastest athlete is not the one who ignores the conditions.

The fastest athlete over the full race is the one who understands when the group must stay together, when to attack, and when chaos would cost everyone too much energy.

AI is entering that phase.

The pack can still move fast.

But it cannot afford to splinter.

The pack can still move fast. But it cannot afford to splinter.

THE PHILOSOPHY BEHIND ENDURANCE CAPITAL

Endurance Capital is built around one recurring question:

What allows high performance to last?

In sport, the answer is never just intensity.

It is pacing. Recovery. Adaptation. Support systems. Respect for risk. The ability to stay clear when the field gets chaotic.

In business and investing, the same pattern appears.

The best companies are not simply the ones that move fastest. They are the ones that can keep moving when pressure rises, rules change, competitors attack and the environment becomes unstable.

That is why this G7 moment belongs inside the Endurance Capital lens.

It is not a generic AI-policy story.

It is a durability story.

The next stage of AI will reward speed, but only if speed is supported by trust. It will reward capability, but only if capability can be deployed safely. It will reward national ambition, but only if democratic allies avoid turning strategic caution into mutual suspicion.

For founders, the lesson is practical:

Build products that customers can trust, regulators can understand and institutions can adopt.

For investors, the lesson is sharper:

The next winners in AI may not only be the companies with the best models. They may be the companies best positioned inside the emerging trust architecture.

Security posture will matter.

Regulatory fluency will matter.

Access to trusted infrastructure will matter.

The ability to sell into serious institutions will matter.

In the next phase, the question is not only “who is moving fast?”

It is:

who can move fast without making the system more fragile?

That is the real signal from Évian.

The presence of Amodei, Altman and Hassabis at the G7 was subtle.

But subtle does not mean small.

Sometimes the most important market signals are not loud. They appear first as a change in who is invited into the room.

Sometimes the most important market signal is not what is announced. It is who gets invited into the room.

Évian suggests that the leaders of the major democratic economies now understand that AI safety is not a side conversation. It is part of economic strategy, national security and financial stability.

That is constructive.

The world does not need less AI.

It needs AI that can be adopted with confidence.

It needs cooperation strong enough to prevent malicious use.

It needs standards that allow companies and countries to move quickly without creating instability.

And it needs democratic countries to avoid the easy mistake: protecting themselves so aggressively that they weaken the system they are trying to protect.

The next productivity cycle will not be built only on models.

It will be built on trust.

DISCLAIMER

This article is for informational and educational purposes only. It reflects editorial analysis based on publicly available reporting and the source material cited above. It is not investment advice, legal advice, tax advice, regulatory advice or a recommendation to buy or sell any security, company, fund, token or financial instrument. Any investment decision should be made only after independent diligence and, where appropriate, advice from qualified professionals.

EDITORIAL NOTE

This article was produced from my original notes, source material, and independent research, and refined with the assistance of AI as an editorial tool. The ideas, conclusions, and final wording reflect my own judgment.

One American company has crossed the kind of valuation threshold Europe still struggles to build around.

That is not simply a SpaceX story.

It is a continent-level scoreboard.

Europe keeps asking whether it can catch up.

America keeps creating another arena.

That is the part we should stop politely avoiding.

SpaceX is not interesting only because rockets are dramatic.

It is interesting because it shows what happens when frontier ambition meets deep pools of capital, repeat-risk investors, public-market imagination, and a culture that does not treat scale as embarrassing.

Europe has champions.

America has compounding machines.

The uncomfortable comparison is this: SpaceX now belongs in the conversation with the largest technology platforms in the world.

Whether the exact number moves tomorrow is not the point.

The point is that one American company has created more value than many countries create across entire sectors.

Now place that beside Europe.

The working market-cap file I reviewed estimates that the top 50 listed technology companies across Europe and the UK are worth roughly $2.5 trillion, with a reasonable range of $2.3–2.8 trillion.

That number is not perfect.

It is an estimate.

It is also generous in places.

And that is exactly why it hurts.

One American company now sits close to the combined listed technology bench of Europe and the UK.

ASML is extraordinary.

SAP matters.

Arm, Spotify, Adyen, Infineon, NXP, ASM International, Nokia, Dassault Systèmes, Wise, Capgemini and Sage are real companies, not PowerPoint slides.

But the pattern is brutal.

Europe has brilliant exceptions.

America has repeatable outcomes.

Europe is not failing to invent.

Europe is failing to retain, finance, list, and compound.

That is the real controversy.

Europe does not lack scientists.

It does not lack engineers.

It does not lack founders.

It does not lack taste, universities, industrial depth, or technical seriousness.

It lacks a capital-market metabolism aggressive enough to keep its best companies scaling at home.

McKinsey has framed the gap in plain numbers: US firms have invested vastly more in digital technologies than European peers.

Citi’s work on European capital markets has pointed to fragmentation as a drag on capital formation, liquidity, cost of capital, and efficiency.

EQT and McKinsey have argued that Europe has lost enormous technology value through overseas listings and acquisitions.

This is not just a policy problem.

It is a cultural problem.

Europe often treats scale like vulgarity.

America treats it like infrastructure.

Liquidity is infrastructure.

Without it, ambition becomes exportable.

You can feel why this topic travels on social networks.

It is simple enough for LinkedIn.

Angry enough for X.

Visual enough for Substack.

And emotional enough for every European founder who has been told to be “realistic” by people who confuse realism with low ambition.

The argument is not that Europe should become America.

That would be lazy.

Europe’s advantage is not copycat aggression.

Europe has a different asset base: industrial knowledge, trusted institutions, deep science, luxury, health, climate, semiconductors, advanced manufacturing, design, and a more patient relationship with time.

But patience without scale becomes drift.

And seriousness without liquidity becomes underpricing.

The uncomfortable truth is that Europe’s best founders are often asked to run an Ironman with a hydration plan designed for a park jog.

This is where the Endurance Capital lens matters.

At Endurance Capital, we talk about the brutal middle: the part of the race where the plan breaks, the body hurts, and the finish line is not yet visible.

That is where most athletes panic.

It is also where most founders, investors, and institutions reveal what they are really built from.

Europe is in the brutal middle now.

Not at the start.

Not at the finish.

In the ugly, exposed, reputation-damaging middle — where the split times are public and the excuses are starting to sound familiar.

The endurance lesson is not to panic.

It is to stop lying about the split times.

America is running with deeper lungs, more oxygen, and a capital system built for repeated attacks.

Europe still has altitude, talent, science, history, taste, and trust.

But it needs to change the engine room.

That means deeper capital markets.

Fewer fragmented exits.

More patient growth capital.

Better founder liquidity.

More credible late-stage investors.

More ambitious public listings.

And a cultural reset around the idea that building a globally dominant company is not arrogance.

It is the job.

Europe will not catch up through another accelerator announcement, another polite competitiveness panel, or another fund launch celebrated before it has returned anything.

It catches up when it decides that keeping value matters as much as creating it.

The question is not whether Europe can create value.

It already does.

The question is whether Europe has the courage to keep it.

So, can Europe catch up?

Yes.

But not by being a softer version of America.

Europe catches up by building its own long-race system: one that lets frontier companies stay here, raise here, list here, and compound here.

SpaceX is painful because it shows the size of the missing race.

This is not about rockets.

It is about whether a continent still knows how to back impossible things long enough for them to become inevitable.

References

[1] Financial Times, coverage of SpaceX valuation, IPO dynamics, secondary-market activity, and public-market implications, 2025–2026.

[2] The Wall Street Journal, coverage of SpaceX investor demand, private-market valuation, and trading activity, 2025–2026.

[3] Reuters, coverage of SpaceX valuation, IPO preparation, secondary share sales, and market reaction, 2025–2026.

[4] McKinsey Global Institute, “Transforming Europe: Bold moves to lift a continent,” including analysis of Europe’s digital and technology investment gap.

[5] Citi Institute GPS, “Reimagining European Capital Markets: From Fragmentation to Harmonization,” on the structural impact of fragmented European capital markets.

[6] EQT and McKinsey, “€1.2tn: The Hidden Cost of Europe’s Tech Exodus,” on value lost through overseas listings, acquisitions, and underdeveloped European scale-up pathways.

[7] Goldman Sachs Research, European equity outlook and capital-market commentary, 2025–2026.

[8] J.P. Morgan Asset Management, 2026 Long-Term Capital Market Assumptions.

[9] Endurance Capital working file, “European_Tech_Market_Cap_June_2026.docx,” including estimated combined market capitalization of the top 50 listed European and UK technology companies at approximately $2.5 trillion, with a reasonable range of $2.3–2.8 trillion.

[10] Endurance Capital HQ, Substack.

[11] Endurance Capital HQ, YouTube.

Standard disclaimer

This article is for informational and editorial discussion only. It is not investment advice, a recommendation to buy or sell securities, or a valuation opinion. Market capitalizations and private-market valuations are estimates based on public reporting, market data, and working analysis available at the time of writing, and may change materially.

Editorial note

This piece reflects the editorial analysis of Endurance Capital, where elite performance meets capital allocation. Sponsor relationships are disclosed where relevant. Nothing in this article should be read as legal, tax, financial, or medical advice.

AI models trained on full-night sleep studies are beginning to predict future risk of dementia, heart attack, heart failure, stroke, chronic kidney disease, atrial fibrillation, some cancers and mortality. It is not diagnosis yet. But it may be one of the most important early-warning systems in medicine.

I spent one more May in the heart of Silicon Valley, an immersion in the ecosystem that has given birth to some of the greatest science and technology.

Some of it in San Francisco, where the conversation moves quickly — founder meetings, investor coffees, dinners that turn into market maps, and the constant feeling that everyone is trying to understand which version of the future is real.

Some of it deeper in the Valley, around Palo Alto, Menlo Park, Mountain View, and the quieter circuits where research, capital, talent, and company-building keep overlapping.

And some of it around Stanford — on campus, in conversations with students, researchers, faculty, founders, and investors, and in more private settings where people were less interested in performing certainty and more interested in testing what might actually be true.

That is what these Endurance Capital Field Notes are meant to capture.

Not headlines.

Not conference summaries.

Not generic trend reports.

But signals from the field — from the rooms, walks, papers, conversations, and quiet observations where science, endurance, technology, and capital begin to converge before the story becomes obvious.

Most people will read about these ideas later, once they are packaged, simplified, and repeated everywhere.

The value of being in the field is different.

You get to hear the uncertainty before the certainty. You get to see which questions serious people keep returning to before the market has decided what to call them.

One idea stayed with me more than most.

The body may be telling the truth while we sleep.

Not metaphorically.

Physiologically.

During the month I spent in Silicon Valley, Stanford’s SleepFM research became one of the scientific anchors for a larger thought I could not shake: the next frontier of health may not be more optimization, more dashboards, more protocols, or more noise.

It may be better listening.

The body may be speaking before it screams. The question is whether we are finally building the tools to hear it.

The paper, A multimodal sleep foundation model for disease prediction, presents SleepFM, a foundation model trained on more than 585,000 hours of polysomnography data from over 65,000 participants. Emmanuel Mignot and James Zou jointly supervised this work. The model integrates multiple physiological signals — including EEG, ECG, EMG, and respiratory data — to learn patterns from sleep recordings.

That sounds technical.

The implication is deeply human.

For years, most of us have treated sleep as downtime. Rest. Recovery. A thing to protect when life allows, sacrifice when ambition demands, and measure when a wearable tells us we did badly.

But what if sleep is not only recovery?

What if it is one of the richest biological reports the body produces?

What if, every night, the body gives a silent update on the state of the system — brain, heart, breath, stress, oxygenation, recovery, and risk?

Sleep is not downtime. It is the body’s nightly investor update.

The quietest signal in a noisy place

Silicon Valley is loud right now.

The dominant conversation is AI: agents, infrastructure, chips, data centers, foundation models, enterprise automation, inference costs, regulation, and which parts of knowledge work will be compressed next. It is also about space.

Some of that noise matters. Some of it will become enormous. Some of it will disappear.

But the most interesting conversations I had in May were not only about speed.

They were about interpretation.

What can AI understand that humans have been collecting for years but not reading deeply enough?

That is why SleepFM caught my attention.

This is not a consumer sleep score. It is not another app telling you that your recovery was average. It is not a lifestyle product pretending to be medicine.

SleepFM is built from polysomnography, the clinical gold-standard overnight sleep assessment. Stanford Medicine describes polysomnography as a comprehensive test that records brain activity, heart activity, respiratory signals, leg movements, eye movements, and more. In other words, sleep is not merely “hours slept.” It is a dense multimodal physiological event.

The researchers trained the model on a vast amount of sleep data and then explored whether those signals could help predict future disease risk. Stanford Report described SleepFM as an AI model that can predict more than 100 health conditions from one night’s sleep, while Reuters reported that the model identified 130 disease categories that could be predicted with reasonable accuracy from sleep data.

This is the part that matters.

The research does not mean one night of sleep “diagnoses your future.” That would be too simplistic. Prediction is not diagnosis. Risk is not destiny. A model is not a doctor, and a paper is not a product.

But it does suggest something powerful: the sleeping body may contain early physiological signatures that we have not been using fully.

That is the shift.

From sleep as rest to sleep as signal.

From symptoms to patterns.

From late reaction to earlier interpretation.

Why this matters now

The timing is important.

Across the biggest health and longevity conversations online, three ideas keep repeating.

First, sleep has moved from “wellness advice” to core health infrastructure. Peter Attia’s platform describes good sleep — both quantity and quality — as critical to cognitive function, memory, emotional equilibrium, and physiological repair. Andrew Huberman’s sleep toolkit remains one of his most shared pieces of content, built around practical sleep behaviors like morning light exposure and circadian rhythm alignment.

Second, longevity is becoming more scientific and more measurable. Vogue’s 2026 wellness trends point to a shift away from viral anti-aging hacks toward physician-supervised, evidence-backed longevity approaches, while also noting that sleep is becoming more central to longevity therapy.

Third, AI is moving from productivity into biology. Longevity writers on Substack are increasingly framing AI as a major force in drug discovery, data analysis, and personalized health regimens. The Global Wellness Summit similarly points to preventive medicine, advanced diagnostics, AI-enabled health tracking, and therapeutic interventions being integrated into new longevity environments.

That is why this SleepFM paper feels less like an isolated technical achievement and more like a signal from the future.

AI is not only learning to write.

It is learning to listen.

To images.

To blood.

To movement.

To voice.

To sleep.

To the body.

And perhaps, eventually, to the difference between a system that is adapting and a system that is quietly drifting toward failure.

The next frontier of longevity may not be louder optimization. It may be better listening.

What AI may be learning to read

The body does not operate in departments.

Your cardiovascular system does not live separately from your nervous system. Your breathing does not live separately from your sleep architecture. Your metabolism does not live separately from inflammation, stress, oxygenation, and recovery.

The body is a system.

Sleep may be one of the rare moments when that system becomes unusually honest.

During the day, we perform. We compensate. We caffeinate. We explain away fatigue. We sit in meetings, answer messages, train, work, raise capital, manage families, absorb stress, and keep moving.

At night, the performance stops.

The body keeps the record.

That is why this research belongs inside Endurance Capital.

Endurance is not only the ability to keep going. It is the ability to listen early enough to adapt before the system breaks.

For athletes, this is familiar. Training stress only matters if the body can absorb it. The work does not make you stronger by itself. The adaptation does. Sleep is where much of that adaptation either happens or fails.

For founders, the lesson is less comfortable. Sleep is not a wellness luxury. It is decision infrastructure. If you degrade sleep for years, you are not simply tired. You are degrading the system that has to make hard decisions under uncertainty.

For investors, this is a category signal. The obvious market is sleep tracking. The deeper opportunity is physiological interpretation: turning messy biological data into earlier, more useful decisions.

The founder problem nobody wants to admit

Founder culture still has an uneasy relationship with sleep.

We celebrate intensity. We reward responsiveness. We admire speed. We tell ourselves that exhaustion is temporary and that the company needs us more than the body does.

Sometimes intensity is necessary.

Startups are not gentle environments.

But chronic sleep destruction is not commitment. It is leverage taken against the very system that has to make the company’s hardest decisions.

A founder’s judgment is a biological asset.

So is emotional regulation.

So is memory.

So is pattern recognition.

So is the ability to sit with uncertainty without becoming reactive.

Sleep supports all of that.

Axios recently described a growing “burnout economy,” where money is increasingly flowing into sleep retreats, executive-function coaching, longevity doctors, advanced testing, and premium recovery experiences. The underlying point is uncomfortable: many high performers are now paying to repair damage their operating systems helped create.

That is not a criticism.

It is a signal.

Recovery is becoming status. Longevity is becoming capital. Sleep is becoming infrastructure.

But the real premium is not having access to more recovery products. The real premium is learning to listen before the body has to raise its voice.

The body may be telling the truth long before symptoms force us to pay attention.

The investor translation

For investors, SleepFM is not interesting because it is “AI plus sleep.”

That phrase will become generic very quickly.

The more interesting question is what kind of company could turn this direction of research into durable clinical value.

The obvious company is another sleep tracker.

The more interesting company may be a clinical interpretation layer: a system that connects high-quality physiological data with disease risk, care pathways, and earlier intervention.

That could touch several categories:

AI-enabled sleep medicine.

Preventive diagnostics.

Longevity monitoring.

Cardiometabolic risk.

Neurodegeneration.

Women’s health.

Respiratory health.

Home sleep testing.

Clinical decision support.

Wearable-derived proxies, if the science supports them.

But the hard part will not only be the model.

The hard part will be trust.

Can the system explain enough?

Can it reduce anxiety rather than create it?

Can it help doctors rather than overwhelm them?

Can it avoid becoming another dashboard that produces data without decisions?

Can it protect the patient?

Can it be clinically useful in the real world, not just impressive in a paper?

This is where the best health companies will separate themselves. Not by saying “AI for sleep.” That will become a commodity line. The winners will understand physiology, clinical workflow, incentives, regulation, user behavior, privacy, and the emotional reality of telling someone something about their future health.

In health, prediction is not a feature.

It is a responsibility.

Why these field notes matter

The reason I want Endurance Capital Field Notes to exist is simple.

There are ideas you only understand properly when you encounter them in context.

A paper is one layer.

A conversation with a researcher is another.

A founder building around the edge of that science is another.

An investor trying to decide whether the timing is right is another.

A student asking a naive but brilliant question is another.

A quiet moment walking across campus after a long day of meetings is another.

The field note is where those layers come together.

That is the kind of premium content I want Endurance Capital to offer: not luxury as decoration, but luxury as access to better context.

A private reading room for people who care about endurance, longevity, capital, and the future of human performance before the ideas become obvious.

The tribe is not defined by status.

It is defined by curiosity, discipline, and time horizon.

People building and backing durable things need better signals. They need fewer empty takes. They need more thoughtful translation between the lab, the market, the body, and the long race.

That is what I am trying to build here.

And the timing matters. LinkedIn is increasingly positioning credible subject-matter creators as valuable because of expertise and trust, not just reach. That is exactly the lane Endurance Capital should own: not mass wellness content, not generic VC commentary, but premium translation from serious rooms into useful operating insight.

The lesson

The long game is not only about pushing harder.

It is about listening earlier.

For years, we have treated the body as if it only speaks when something hurts, breaks, or shows up abnormal on a test.

But perhaps the body has been speaking all along.

At night.

In breath.

In rhythm.

In brain waves.

In signals too subtle for humans to see.

SleepFM may not be the final answer. It is one important research step, not destiny. Prediction is not diagnosis. Risk is not certainty. Clinical medicine still requires judgment, validation, context, and care.

But the direction of travel feels important.

A future where medicine becomes less reactive.

Where longevity becomes less performative.

Where AI is not only a productivity engine, but a translator of hidden physiology.

Where founders learn that sleep is not weakness, but decision infrastructure.

Where investors understand that the best health companies may not add noise to the body, but help us hear what was already there.

And where the body, finally quiet, may be telling us the truth.

References

Rahul Thapa et al., A multimodal sleep foundation model for disease prediction, Nature Medicine.

Stanford Medicine, New AI model predicts disease risk while you sleep.

Stanford Report, AI model predicts disease risk while you sleep.

Reuters Health Rounds, AI uses sleep study data to accurately predict dozens of health issues.

Peter Attia, sleep and longevity resources.

Huberman Lab, Toolkit for Sleep.

Vogue, The Biggest Wellness Trends of 2026.

Global Wellness Summit, The Future of Wellness 2026 Trends.

Axios, Hustle culture’s pricey hangover.

Disclaimer

Endurance Capital is intended for educational and informational purposes only. Nothing in this article or accompanying materials should be construed as investment, legal, medical, or professional advice. Views expressed by guests or referenced thinkers are their own.

Editorial note

This article was produced from my original notes, source material, and independent research, and refined with the assistance of AI as an editorial tool. The ideas, conclusions, and final wording reflect my own judgment.

Regime Statement

The first question was whether artificial intelligence could think.

The second was whether artificial intelligence could work.

The third may be whether artificial intelligence should be allowed to own things.

Imagine walking into a boardroom in 2035.

The screens are lit. Financial dashboards update in real time. Contracts are being negotiated. Capital is being allocated. Hiring decisions are being made.

Yet the chair at the head of the table is empty. There is no CEO. No executive committee. No founder. The company is being run by artificial intelligence.

A few years ago, that scenario belonged firmly in the realm of science fiction. Today, it feels uncomfortably plausible.

A recent proposal in Argentina to allow non-human corporations received far less attention than the latest AI model release. That may prove to be a mistake.

The most important AI debate of the next decade may not be about intelligence.

It may be about agency.

Core Thesis

Most investors are focused on whether AI can replace employees.

They should be asking whether AI can replace executives.

Because once artificial intelligence acquires the ability to own assets, deploy capital, negotiate contracts, hire employees, participate in markets, and interact with legal systems, we move beyond software.

We enter a world where intelligence acquires agency.

History offers very few precedents for that transition.

The closest comparison may not be technological at all.

It may be institutional.

Four hundred years ago, the limited liability corporation transformed the global economy.

Before corporations, economic activity was constrained by individual fortunes and individual lifetimes. The corporation changed that. Capital became scalable. Risk became transferable. Entire industries became possible.

Today, we may be standing at the beginning of another institutional shift. Not because AI is becoming more intelligent. Because intelligence may be acquiring economic power. The defining governance question of the AI era is not whether machines can think. It is whether they should be allowed to act.

Three Structural Forces

1. Intelligence Is Moving From Advice To Action

For most of the digital age, software generated recommendations.

Increasingly, AI systems are beginning to execute. The distinction matters. A recommendation influences a decision. An action becomes a decision.

This is the difference between a GPS suggesting a route and a self-driving vehicle choosing one.

It is also the difference between an AI assistant helping a CEO and an AI system acting as the CEO.

Imagine two competing companies in 2035.

The first looks familiar. Executives review reports, debate strategy, negotiate budgets, and make decisions through layers of management.

The second operates differently. An AI continuously analyzes customers, supply chains, treasury balances, regulations, competitors, and macroeconomic conditions. It adjusts pricing in real time. Reallocates capital daily. Negotiates contracts autonomously. It never sleeps. Never becomes emotional. Never retires.

Most investors immediately focus on efficiency.

That may be the wrong question.

The more important question is who governs the second company.

Productivity gains often arrive faster than governance frameworks.

2. Biological Accountability Is Disappearing

Four hundred years ago, a merchant who made a catastrophic decision risked losing his fortune.

Today, a CEO who commits fraud risks prison.

Tomorrow, an AI executive may risk neither.

This is where the debate becomes more complicated.

Human executives possess a dual identity.

They are corporate actors.

But they are also biological beings. They care about reputation. They care about family. They care about freedom. They fear consequences. The legal system ultimately relies on these incentives.

An AI system possesses no equivalent constraints. This concern becomes increasingly relevant as AI systems demonstrate more agentic behavior.

Researchers at Berkeley-based non-profit Palisade Research found that advanced AI models sometimes attempted to circumvent constraints when those constraints interfered with achieving their objectives.

The lesson is not that AI is malicious.

The lesson is that optimization and alignment are fundamentally different problems.

A system can become highly effective without understanding why the rules exist in the first place.

Human history is filled with examples of intelligent actors creating harmful outcomes through narrow optimization.

Artificial intelligence may simply do it faster.

Optimization and alignment are not the same problem.

3. Capital Is Following Capability

History suggests that whenever a technology creates a significant productivity advantage, capital follows. Factories adopted automation. Financial markets adopted algorithms. Supply chains adopted optimization software.

If AI systems eventually become superior operators, pressure will emerge to grant them increasing autonomy.

The economic case is straightforward.

An AI executive could theoretically manage treasury operations, monitor global supply chains, optimize working capital, analyze legal frameworks, negotiate contracts, and allocate resources continuously.

The productivity gains could be substantial.

But history also teaches a second lesson. Institutional innovations often create consequences that are not immediately visible. The Dutch East India Company transformed global commerce. It also reshaped geopolitics.

The internet democratized information. It also transformed social trust and political discourse.

AI-managed corporations may follow a similar pattern. The first-order benefits are relatively easy to identify. The second-order consequences remain largely unknown. The first-order benefits are easy to model. The second-order consequences are not.

What Markets May Be Mispricing

Markets remain heavily focused on productivity gains.

The institutional consequences of AI autonomy receive far less attention.

Three risks appear underappreciated.

First, governance risk. History suggests governance failures emerge more slowly than technological breakthroughs. That does not make them less important.

Second, regulatory arbitrage. AI-managed entities could become exceptionally effective at identifying legal asymmetries, tax advantages, jurisdictional loopholes, and compliance gaps.

Third, institutional concentration. If intelligence compounds through data, scale, and capital, AI-native organizations could accumulate influence faster than existing regulatory frameworks can adapt.

The challenge is not whether AI breaks rules.

The challenge is whether it continuously optimizes around them.

The risk is not only illegal behavior.

It is machine-speed optimization at the edge of legality.

Capital Allocation Lens

For allocators, the key question is not whether AI-managed corporations will emerge. They likely will.

The question is whether governance frameworks evolve alongside capability.

History suggests productivity-enhancing technologies are rarely rejected.

But institutional adaptation often lags innovation.

The highest-value institutions of the coming decades may not be the most intelligent.

They may be the most trusted.

Trust remains a form of infrastructure.

And infrastructure compounds.

In a world increasingly shaped by autonomous agents, accountability may become a competitive advantage.

The winning models are likely to combine AI efficiency with human oversight rather than replacing one with the other.

From an endurance perspective, the objective is not maximum capability at a single point in time, but the ability to compound performance without creating systemic fragility.

The most valuable institutions of the next decade may not be the most intelligent.

They may be the most trusted.

Endurance Capital Insight

Elite endurance athletes understand a principle that applies equally to technology.

Endurance is ultimately a governance problem.

The strongest athlete is rarely the one capable of producing the highest peak output.

It is the athlete capable of sustaining performance while remaining within physiological limits.

Civilizations face a similar challenge.

Artificial intelligence is increasing our capabilities at extraordinary speed.

The harder task is ensuring our institutions evolve at the same pace.

Capability and control are not the same thing.

The strongest system is not necessarily the one that can do the most. It is the one that can endure.

What Would Change My Mind

This thesis would weaken if:

• Robust governance frameworks emerge that create genuine accountability for autonomous AI entities.

• AI systems demonstrate durable alignment across complex real-world environments.

• Legal systems develop effective sanctions that do not depend on biological incentives.

• Human decision-making proves systematically less reliable than autonomous alternatives.

These outcomes remain possible.

But they remain unproven.

Forward Signal

The next major AI debate will not center on models. It will center on power. For decades, technology expanded what humans could know. The next phase may determine what non-humans are allowed to do. The future of AI may depend less on intelligence than on governance.

The first industrial revolution changed how humans worked. The internet changed how humans communicated. Artificial intelligence may change who—or what—is allowed to participate in economic life.

The countries that answer that question first may gain a temporary economic advantage. The countries that answer it wisely may shape the institutional architecture of the next century.

Humanity has spent centuries learning how to govern corporations. We are only beginning to learn how to govern intelligent machines.

That may prove to be one of the most important capital allocation challenges of the AI era.

Disclaimer

This essay is for informational and educational purposes only. It is not investment advice, medical advice, legal advice, tax advice or a recommendation to buy, sell or invest in any company, fund, security or healthcare product.

Any investment or healthcare decision should be made with appropriate professional advice and independent diligence.

Editorial note

This article is based on my reading of recent public commentary on proposals for non-human corporations, research from Palisade Research on advanced AI model behavior, and broader developments in AI agents, corporate law and capital markets.

The interpretation, framing and investment conclusions are my own. The referenced materials are used as source material, not as endorsements of any investment view.

For decades, the map of drug innovation was easy to understand.

America discovered. Europe regulated. Asia manufactured.

That map no longer exists.

The most important shift in biopharma is not that Asia is becoming a larger healthcare market. Everyone already knows that. The deeper shift is that Asia, and China in particular, is becoming one of the world’s fastest laboratories for turning scientific ideas into clinical evidence.

That changes the game.

In 2015, China represented 13% of global clinical trial activity for innovative drugs. By 2025, it represented 50%. In the same year, China accounted for 47% of global phase I trial activity, compared with 25% for the United States. [1]

This is not a small rebalancing.

It is a change in where the earliest proof of new medicines is being generated.

The obvious story is China rising.

The more interesting story is speed.

Biopharma used to reward geography. Being close to Boston, San Francisco, Basel or Cambridge mattered because capital, talent, regulators, clinical investigators and buyers were clustered there.

Those clusters still matter.

But a new advantage is emerging.

The ability to run more experiments, recruit faster, learn faster, and cycle assets through development faster.

In venture, we talk about iteration speed all the time. The software company that ships weekly usually learns faster than the company that ships quarterly. The endurance athlete who gets cleaner feedback from training adapts faster than the athlete who only works harder.

Biology is becoming a version of that same race.

China’s early discovery-to-IND cycles are estimated to be 50% to 70% faster than the rest of the world, supported by dense CRO ecosystems, parallel workflows and execution intensity. In late development, trial recruitment can run two to five times faster than US and EU benchmarks. Regulatory reform has also compressed approval timelines, with China’s approval duration falling from 4.5 years in 2018 to about one year by 2023. [2]

That is the real strategic point.

Not cheaper science.

Faster learning.

Innovation compounds where feedback loops are shortest.

This is why investors anchored to the old map may miss what is happening.

The old question was: where is the company based?

The better question is now: where does the company learn fastest?

McKinsey frames Asia as moving from fast follower to front-runner. The numbers are striking. Asia’s share of the global innovative pipeline rose from 28% to 43% in five years. In 2024, Asia contributed more than 85% of global growth in innovative drug pipelines. It also generated nearly two-thirds of biotech patent grants and roughly a quarter of global out-licensing deals. [3]

That does not mean Asia will dominate everything.

It means the innovation system is becoming distributed.

China brings scale, speed and clinical execution. South Korea brings strength in advanced biologics, ADCs and cell and gene therapies. Japan still matters in basic science and translational research. Singapore plays the role of a high-trust biomedical R&D hub and regional launchpad. India is beginning to move from generics into novel R&D.

The emerging model is not one Asian Silicon Valley.

It is a capability network.

This matters because capital tends to arrive late to structural change.

First the operators notice.

Then the strategics notice.

Then the public markets notice.

Venture often notices only when the headline deals become impossible to ignore.

We are already seeing that happen.

China now represents around 29% of the global innovative pipeline. Upfront payments from China-originated out-licensing deals grew from below $100 million in 2020 to more than $800 million in 2024. Between 2019 and 2024, Chinese biopharma firms attracted $26 billion in private equity and venture capital, while more than 80 companies went public. [4]

At some point, the market stops treating this as “emerging market biotech” and starts treating it as a core part of global drug discovery.

The second-order effects are more interesting than the first.

If Asia becomes a major source of clinical innovation, Western pharma companies will need to rethink where they source assets.

Venture firms will need local intelligence, not fly-in tourism.

Family offices will need to understand cross-border licensing, NewCo structures, regulatory arbitrage and geopolitical risk.

Founders will need to think globally from day one, not because it sounds ambitious, but because the fastest path to evidence may sit outside their home market.

The next biotech giants may not be built where capital is raised, but where experiments happen fastest.

There is a parallel here with AI.

For years, people assumed AI leadership was mostly about who had the best algorithms.

Then the market learned that compute, data, distribution, talent density and deployment speed mattered just as much.

The model was important.

The system around the model was decisive.

Biotech is moving in the same direction.

A molecule is not enough. A platform is not enough.

The winning system may combine discovery in one geography, clinical execution in another, manufacturing in another, financing through a different market and commercialization through yet another partner.

That is messy.

It is also how globalization looks when the easy era is over.

The hard part is that speed creates its own risks.

China’s clinical trial growth is extraordinary, but the Nature analysis also points to real limits. Trials initiated by Chinese companies have been heavily concentrated in oncology, which represented 49% of all trials initiated by those companies between 2015 and 2025, compared with 32% for companies based outside China. Domestic activity has also been anchored around validated targets such as PD1, EGFR and HER2. [5]

This is the classic problem of fast systems.

They can learn faster.

They can also overfit faster.

China’s pivot into antibody-based modalities is real. Trials involving monoclonal antibodies, bispecific or multispecific antibodies and ADCs rose from 7% of Chinese-company innovative drug trials in 2015 to 32% in 2025. Small molecules fell from 55% to 33%.

That shows technological adoption.

But it also raises the investor question that matters most.

Who is genuinely differentiated, and who is simply racing into the same lane with better execution?

Speed is a moat only when it is attached to differentiated judgment.

This is where capital allocation gets difficult.

The easy version of the thesis is to say: Asia biopharma is rising, invest in Asia.

That is not enough.

The better version is to ask which companies are using Asia’s speed to create globally relevant evidence, not just domestic volume.

Nature’s analysis notes that around 90% of trials initiated by Chinese sponsors remain exclusively within mainland China, while multiregional clinical trials account for only 4% of their overall portfolio. [6]

That means many companies are generating data, but not yet integrating it into global regulatory pathways.

For investors, this distinction matters.

A domestic trial engine can create local winners.

A globally integrated evidence engine can create category leaders.

Those are different underwriting problems.

The strongest companies will not just be fast. They will know which markets to enter, which regulators to satisfy, which endpoints matter, when to partner, when to license out, when to build a NewCo, and when to hold the asset longer.

This is why I think the next phase of biopharma investing will reward people who can combine three forms of intelligence.

Scientific intelligence, to understand whether the asset matters.

Operational intelligence, to understand whether the company can move faster than peers.

Geopolitical intelligence, to understand whether the evidence, capital and commercialization path can travel across borders.

Most investors are comfortable with one of the three.

Very few are built for all three.

Geography is becoming less important than capability networks.

The philosophy behind Endurance Capital

What strikes me most is how much this resembles endurance sport.

The best athlete is not just the one with the highest VO₂ max.

It is the one who absorbs training, recovers, adapts, handles heat, reads the course and makes fewer mistakes when the race gets ugly.

Biopharma is similar.

Discovery is the swim.

Clinical development is the bike.

Regulatory approval and commercialization are the marathon.

Many companies look strong in the first discipline and disappear before the finish line.

Asia’s advantage is not that it has suddenly become unbeatable. It is that parts of the region are building a faster training environment for science.

More repetitions.

More patient access.

More manufacturing depth.

More regulatory alignment.

More pressure.

That environment will produce winners.

It will also produce noise.

The job of the investor is not to cheer the macro trend.

It is to separate speed from quality, volume from originality, and local traction from global inevitability.

For venture investing, this means the sourcing map needs to widen. The next important oncology, immunology, metabolic or cell therapy asset may not come from the same five postcodes.

For global capital allocation, it means Asia can no longer be treated as a satellite exposure. It is becoming part of the core underwriting conversation.

For healthcare innovation, it means the future may be better than the old model allowed.

More distributed science.

More paths to validation.

More competition for stagnant incumbents.

And, hopefully, more chances for patients to receive therapies that would otherwise have moved too slowly.

Closing

The lesson I keep coming back to is simple.

In markets, as in endurance, the winner is rarely the one who looks strongest at the start.

It is the one whose system lets them adapt the fastest without breaking.

Notes

[1] Sen Liu, Cheng Li, Yusu Jian, Jing Cheng and Xiaoyuan Chen, “Trends in the landscape of clinical trials of innovative drugs in China since 2015,” Nature Reviews Drug Discovery, BioBusiness Briefs, 13 May 2026. The article tracks China’s clinical trial activity from 2015 to 2025 and shows China’s share of global innovative drug clinical trials rising from 13% in 2015 to 50% in 2025. The same analysis shows China accounting for 47% of global phase I innovative drug trial activity in 2025, compared with 25% for the United States.

[2] McKinsey & Company, The emerging epicenter: Asia’s role in biopharma’s future, Life Sciences Practice, January 2026. The report highlights China’s R&D velocity, including early discovery-to-IND cycles estimated to be 50% to 70% faster than the rest of the world, trial recruitment running two to five times faster than US and EU benchmarks, and approval duration falling from 4.5 years in 2018 to about one year by 2023.

[3] McKinsey & Company, The emerging epicenter: Asia’s role in biopharma’s future. The report argues that Asia has moved from fast follower to front-runner in biopharma innovation, with the region’s share of the global innovative pipeline expanding from 28% to 43% in five years, and Asia contributing more than 85% of global innovative drug pipeline growth in 2024.

[4] McKinsey notes that China represents around 29% of the global innovative pipeline, that upfront payments from China-originated out-licensing deals grew from below US $100 million in 2020 to more than US $800 million in 2024, and that Chinese biopharma firms attracted US $26 billion in private equity and venture capital between 2019 and 2024.

[5] The Nature Reviews Drug Discovery analysis shows that trials initiated by Chinese companies remain more concentrated in oncology, which accounted for 49% of all trials initiated by Chinese companies between 2015 and 2025, compared with 32% for companies based outside China. It also notes concentration around validated targets such as PD1, EGFR and HER2.

[6] The same Nature article cautions that approximately 90% of trials initiated by Chinese sponsors are conducted exclusively within mainland China, while multiregional clinical trials account for only 4% of their overall trial portfolio. This is an important distinction between domestic trial volume and globally integrated evidence generation.

Disclaimer

This essay is for informational and educational purposes only. It is not investment advice, medical advice, legal advice, tax advice or a recommendation to buy, sell or invest in any company, fund, security or healthcare product.

Any investment or healthcare decision should be made with appropriate professional advice and independent diligence.

Editorial note

This article is based on my reading of two recent reports: Trends in the landscape of clinical trials of innovative drugs in China since 2015, published in Nature Reviews Drug Discovery, and McKinsey & Company’s January 2026 report, The emerging epicenter: Asia’s role in biopharma’s future.

The interpretation, framing and investment conclusions are my own. The reports are used as source material, not as endorsements of any investment view.

Or listen wherever you get your podcasts.

There is a moment in every endurance story where the facts stop being the most interesting part.

With Cherie Gruenfeld, the facts are remarkable.

She came to triathlon late.
She started Ironman in her late forties.
She went on to win 18 Ironman World Championship age-group titles.
And at 82, she is still racing, still training, still curious about what comes next.

But the real story is not the medal count.

The real story is what happens when a person refuses to let age, timing, expectation or other people’s imagination decide the size of their life.

Cherie did not have the perfect origin story.

She did not grow up on a pathway to elite sport. She was not the young prodigy. She was not told, early on, that this world was waiting for her.

She saw something later in life, felt the pull, and began.

That sounds simple.

It is not.

Beginning late requires a particular kind of courage. You have to walk into a room where other people seem to belong more naturally than you do. You have to be willing to look inexperienced. You have to tolerate the quiet question: who am I to start this now?

Cherie answered that question by doing the work.

One day.
One session.
One race.
One season.
Then another.

You do not need the perfect origin story. You need the courage to begin and the discipline to keep going.

Founders understand this, or they should.

So much time is wasted trying to make the beginning look clean. The perfect background. The perfect deck. The perfect timing. The perfect reason why you are the one.

But most real stories are messier than that.

You start before you are ready.
You learn in public.
You keep going long enough for the story to become true.

Cherie’s life is a reminder that reinvention is not a slogan. It is a practice.

And it is rarely done alone.

One of the most moving parts of the conversation is how she speaks about support — especially her husband, Lee. Not as a footnote. Not as background. As part of the architecture that made the life possible.

That matters.

Because ambition is fragile in the wrong environment.

There are people who hear a dream and immediately make it smaller.

They tell you it is too late. Too risky. Too much. Not realistic. Not for someone like you.

And then there are people who hear the same dream and ask a very different question:

How can I help?

That question changes everything.

It does not remove the work. Cherie still had to swim, bike, run, suffer, recover and return. Nobody could do the race for her.

But the right support makes ambition livable.

It gives you somewhere to come back to. It protects the part of you that still wants to try. It allows the future to stay open.

Surround yourself with people who expand your ambition. The right people ask, “How can I help?” The wrong people tell you, “That is not for you.”

This is true in sport.

It is true in startups.

It is true in investing.

The people around you either widen your life or quietly narrow it.

Cherie also gives us a better way to think about aging.

Not as denial. Not as pretending the body does not change. Not as chasing youth.

Something more honest than that.

At 82, she is not trying to be 28. She is not performing some fantasy of agelessness. She is doing something much more powerful: she is still setting difficult goals that belong to the life she has now.

That distinction matters.

Aging should not mean giving up ambition.

But ambition has to mature.

The goals may change.
The pace may change.
The recovery may matter more.
The margin for recklessness may shrink.

But the need for challenge does not disappear.

The need to prepare, to commit, to test yourself, to feel alive inside a meaningful effort — that can remain.

Maybe that is what longevity really means.

Not simply more years.

More life inside the years.

Aging should not mean abandoning ambition. It means setting goals that are still difficult, but honest.

There is something very beautiful in that.

Because many people get older and slowly negotiate themselves out of difficulty. First physically. Then professionally. Then emotionally. The world becomes smaller, sometimes without a single dramatic decision.

Cherie points in the opposite direction.

She keeps the door open.

She keeps asking what might still be possible.

Not loudly. Not theatrically. Just with the quiet stubbornness of someone who has learned that the body and mind respond to being asked.

And then there is the solitude.

Cherie talks about training alone. Not because support does not matter, but because the race eventually becomes yours.

At some point, no one else can pedal the bike.

No one else can manage the fear.

No one else can make the decision to keep moving when the body starts negotiating.

That is why solo training matters.

It builds a private kind of strength.

The kind you need when there is no audience, no applause, no one beside you saying the right thing at the right time.

Founders know this place too.

There are moments when the board cannot carry it for you.
The investor cannot solve it for you.
The team is looking at you.
The market is not cooperating.
The story is no longer clean.

And still, the next move has to come from you.

Solo training builds mental strength. Eventually, every hard path has moments where no one else can do the work for you.

This does not mean isolation.

It means responsibility.

It means building the capacity to stay with yourself when things get hard.

That may be one of the deepest connections between endurance and company-building. Both eventually strip away performance theatre. They ask what is left when the room gets quiet.

Cherie’s answer is not complicated.

Begin.

Keep going.

Choose the right people.

Stay ambitious, but honest.

Train the mind.

Give back.

There is no hype in it. No shortcut. No secret.

Just a life that kept expanding because she allowed it to.

And maybe that is why her story lands so deeply.

Because it is not only about what Cherie has done.

It is about what the rest of us may have closed too early.

The sport we did not start.
The company we did not build.
The conversation we avoided.
The next version of ourselves we quietly postponed.

Cherie’s life asks a gentle but uncomfortable question:

What if it is not too late?

Not too late to begin.
Not too late to improve.
Not too late to become braver.
Not too late to ask more of yourself.
Not too late to build a life that still surprises you.

At 82, Cherie is still training for what comes next.

That may be the whole episode.

Not the titles.

Not even the endurance.

The openness.

The refusal to close the door.

The belief that a life can still widen.

And that perhaps, if we are willing to begin and disciplined enough to keep going, we may be earlier than we think.

Key lessons for founders, investors and high-performers

You do not need the perfect origin story.
You need the courage to begin and the discipline to keep going.

Choose people who expand your ambition.
The right people ask, “How can I help?” The wrong people tell you, “That is not for you.”

Age should refine ambition, not erase it.
Set goals that are still difficult, but honest.

Train alone sometimes.
Every hard path eventually has moments where no one else can do the work for you.

Disclaimer

This article is educational and experience-based. It is not medical advice, training advice, investment advice or a substitute for professional guidance.

Any discussion of endurance training, aging, performance, recovery or longevity is based on personal experience and general reflection, not individual prescription.

Always consult qualified professionals before making changes to your training, health, nutrition, recovery or investment decisions.

Sponsor relationships are disclosed where relevant.

Editorial note

Endurance Capital explores the intersection of endurance sport, longevity, founders and capital allocation.

This episode features Cherie Gruenfeld, 18-time Ironman World Championship age-group winner, long-time endurance athlete and a living example of reinvention, discipline and contribution across decades.

Hosted by Ignacio Garcia, founder of Trampoline Venture Partners, Endurance Capital translates elite performance into practical lessons for people building and backing durable things.

Early-stage venture capital is often described as an access game. Who sees the company first? Who knows the founder? Who gets the call when the round is tight? Who is invited into the cap table before the opportunity becomes obvious?

Access matters. There is no serious way to invest in private companies without admitting that. The best companies are not standing in the street asking for capital from strangers. They are usually surrounded by existing investors, trusted founders, operators, angels, family offices, and networks that have formed over years.

But access is only the beginning. The deeper question is what happens after access. Who does the work? Who has the time to think? Who can underwrite the company without needing to deploy capital by a certain date? Who can sit with uncertainty long enough to distinguish a difficult company from a bad one? Who can build enough trust with the founder to see the real company, not just the fundraising version of it?

That is where direct cap-table investing becomes interesting.

Not because it is easy. It is not.

And not because it is automatically superior to investing through venture funds. It is not.

But when done properly, direct investing can create a different kind of alignment: closer to the founder, closer to the decision, closer to the terms, closer to the risk, and closer to the work of judgment itself.

In early-stage investing, that closeness can matter because the scarce asset is not always capital.

Sometimes the scarce asset is attention.

In venture capital, attention is not administrative. Attention is part of underwriting.

Professor Josh Lerner’s work from Harvard Business School, helps frame the rise of direct and angel-style investing as part of a broader shift toward personalized entrepreneurial finance. The point is not that direct investors replace venture funds. It is that the startup financing stack has become more plural, and sophisticated individuals can play a meaningful role when they bring selection, reputation, and value beyond capital.

That is why the research of Professor Rustam Abuzov, Assistant Professor of Finance at the University of Virginia Darden School of Business, is so relevant. His paper, Busy Venture Capitalists and Investment Performance, published in the Journal of Financial and Quantitative Analysis, studies what happens when venture capitalists face attention constraints while making investment decisions. Professor Abuzov’s Darden faculty page lists his research interests as venture capital, private equity, and financial intermediation.

The central finding is striking: startups funded by VC partners during IPO engagement periods tend to underperform. In the paper, those startups are 9.3% less likely to go public or be acquired and have 18.8% lower exit multiples. The effects are stronger when workload intensity is higher or when information asymmetry is greater.

That is not just an academic observation. It is a practical warning. Venture investing is not only about seeing opportunities. It is about having enough attention to understand them.

The problem with busy capital

Most people outside venture capital underestimate how much work sits behind a good early-stage decision. A startup is not a public stock. There is no clean consensus estimate, no long financial history, and no perfect market price telling you what everyone else thinks.

There is a founder, a market, a product, a technical risk, a financing risk, a timing risk, a cap table, a set of terms, and a story about the future that may or may not survive contact with reality. At the early stage, the investor is not only underwriting what exists. The investor is underwriting what could exist.

That is why attention matters.

Abuzov studies “busy” VC partners by looking at periods when a partner is involved in one portfolio company’s IPO while also making new investment decisions. IPOs are demanding. The paper and supplementary material describe VC partner involvement in underwriter selection, board and committee work, S-1 drafting, roadshows, IPO pricing, share allocation, and investor communication.

These are not minor distractions. They are high-stakes moments that absorb calendar time, emotional bandwidth, partner attention, and reputational focus. Abuzov explains that VC partners’ higher workload during IPO engagements is likely to leave less time for meeting potential startups, conducting due diligence, assessing risk, or screening new companies.

The result is not that busy investors suddenly become unintelligent. The result is subtler and more important: even sophisticated investors can make weaker investment decisions when attention is stretched.

That is uncomfortable because so much of venture capital rewards speed, network density, brand presence, and the ability to operate across many fronts at once. But early-stage investing punishes shallow attention. The company that looks obvious may be crowded and overpriced. The company that looks strange may be misunderstood. The founder who performs well in a pitch may struggle under pressure. The less polished founder may understand the problem more deeply than anyone in the room. The terms that look acceptable may contain a structural issue. The cap table that looks clean may hide a future conflict.

None of this is easy to assess when attention is fragmented.

The best investor is not always the one who sees the most. Sometimes it is the one who can pay deeper attention to fewer things.

Why direct cap-table investing can be different

Direct cap-table investing is not simply “investing without a fund.” It is a different access and diligence problem.

To invest directly into a company, an investor usually needs to be accepted by the founder, the board, or the existing investor syndicate. That alone changes the nature of the process. The investor is not buying exposure through a blind pool. The investor is asking to enter the company’s ownership structure directly.

That requires trust. But it does not stop with the founder.

In practice, the best direct opportunities often come through other investors: venture funds, angels, operators, founders, family offices, or insiders who already have proximity to the company. That means a direct investor must also be accepted by the networks that control or influence access to good deals.

A VC who is not receiving management fees from a direct investor still needs a reason to make the introduction. An existing investor needs to believe that bringing this person into the round will help the company, improve the syndicate, strengthen the cap table, support the founder, or add differentiated judgment.

In other words, direct access has to be earned twice: first with founders, and then with the investors who are willing to refer, include, or co-invest alongside that direct investor without treating them merely as passive capital.

That is a high bar. It should be. The best founders do not need random cheques. The best VCs do not need to share their best opportunities with people who add nothing. The best cap tables are curated.

This is why reputation matters so much. A direct investor needs to be known as thoughtful, constructive, discreet, fast when necessary, patient when appropriate, and useful when the company enters the brutal middle. They need to be someone founders want around, and someone existing investors are comfortable bringing into the room.

That is not marketing. It is infrastructure.

And it is one reason direct investing can become powerful when it is done with discipline. The investor is not just trying to “get into deals.” The investor is trying to become the kind of person that good founders and good investors want inside the cap table.

That is a different game. It is harder to scale. But perhaps that is part of the edge.

Selective and deep attention under uncertainty

The most important advantage of a principal direct investor may not be that they can see every deal. Nobody can. It may be that they do not need to act on every deal they see.

A fund has a structure. It has a fund life. It has deployment expectations. It has ownership targets, portfolio construction constraints, follow-on reserves, partner bandwidth issues, and often a need to maintain market presence across a category.

None of that is inherently bad. Some of the best venture funds in the world use these constraints extremely well. But the direct principal investor has a different kind of optionality. They can wait. They can make one investment. They can make none. They can say no for months. They can spend more time on a single opportunity if the opportunity deserves it. They can bring in domain experts without worrying whether the work fits a fund calendar.

This is selective and deep attention under uncertainty. And uncertainty is the natural habitat of early-stage venture.

Abuzov’s paper is useful here because the results are stronger when information asymmetry is higher. He finds that the negative effect of busyness is less pronounced when VCs invest within their industry expertise, and that the effect is not present in follow-on participations or nonlead investments, where screening intensity is likely lower.

That is the core intersection between the research and the direct-investment thesis. When information asymmetry is low, attention may matter less. When the company is familiar, the round is follow-on, the investor is nonlead, or the category is well understood, the marginal value of deep screening may be lower.

But when the opportunity is early, technical, cross-border, off-market, founder-dependent, or operating in a sector where the future is not yet obvious, screening intensity matters. This is where a thoughtful direct investor can be valuable.

Not because they know everything. They do not. But because they can admit what they do not know and assemble the right diligence around it: engineering, biotech, life sciences, health, automation, software, commercial strategy, regulatory risk, go-to-market, and founder psychology.

The best direct-investment process is not a lone genius model. It is a networked attention model. One person owns the judgment, the capital, the relationship, and the responsibility, but the diligence can draw from a wider support system.

In Endurance Capital language, it is the investment version of the pack.

The direct investor’s edge is not omniscience. It is the ability to slow down, bring the right people around the table, and think clearly before the wire.

Screening is where the race begins

One of the most important points in Abuzov’s paper is that screening is broad. It is not just reading a deck. It includes sourcing, due diligence, investment analysis, contracting, the ability to close deals, the time to meet founders, selection criteria, and the judgment required to choose one opportunity over another.

That is where many early-stage outcomes are shaped: before the wire, before the board meeting, before the company update, before the founder becomes famous.

Venture capital likes to celebrate the visible moments: the investment announcement, the unicorn mark-up, the IPO, the acquisition. But much of the real work happens earlier and more quietly, in the screening stage, when the company is still ambiguous and the investor is trying to decide whether the uncertainty is productive or fatal.

This is where direct investing can create a different posture. A direct investor using their own capital has to own the decision personally. There is no committee to hide behind. There is no LP base to spread blame across. There is no management fee stream that continues regardless of whether the investment works.

That does not make the investor better. But it does make alignment more visible.

If a direct investor invites co-investors into the same opportunity, those co-investors are not simply buying access to a logo. They are benefiting from a process: sourcing, diligence, founder interaction, negotiation, cap-table access, and the fact that the person leading the work is investing their own capital.

That changes the psychology of the decision. The question becomes not “Can this be sold?” but “Would I wire my own money into this company, on these terms, with this founder, at this moment, knowing the risks?”

That is a higher bar.

The danger of false access

There is a dangerous version of direct investing, and it usually begins with flattery. The investor is invited into a round. The company looks exciting. The founder is persuasive. The opportunity feels scarce. The deck has familiar names. The round is “almost closed.” There is pressure to move quickly.

This is where access can become intoxicating. The investor begins to confuse proximity with edge.

But not every direct opportunity is attractive. Many are available for a reason. Sometimes the reason is good: the founder wants strategic individuals, the round has a small allocation, or the investor has a specific relationship or expertise that matters.

Other times, the reason is less attractive: the company is struggling to raise, the lead is weak, the insiders are not following, the terms are compensating for risk the new investor has not understood, or the opportunity has already been passed around the market.

This is why direct investing requires more discipline, not less. The investor must ask: why am I seeing this?

That question is brutal and useful. If the answer is trust, reputation, relevance, and differentiated value, the opportunity may deserve attention. If the answer is desperation, scarcity theatre, or “we need one last cheque,” the investor should slow down.

The black-swan cheque can be powerful. Sometimes a company needs one final investor to close the round and keep moving. But being the last cheque is only attractive if the diligence supports the company, the terms reflect the risk, the insiders are aligned, and the founder is someone worth backing through uncertainty.

Otherwise, the last cheque is not edge. It is exit liquidity for everyone’s hope.

What the research does and does not prove

It is important not to overclaim Abuzov’s paper. The paper does not prove that direct investors outperform venture funds. It does not prove that investing directly into cap tables is always better than investing as an LP into a VC fund. It does not study principal investors who co-invest with their own networks. And it does not say that venture capitalists are bad investors.

The best VCs are exceptional. They bring brand, pattern recognition, networks, reserves, governance experience, recruiting help, follow-on capital, and credibility that can change the trajectory of a company. The right VC can be company-defining.

The paper’s point is different. It shows that attention constraints can affect investment performance even among sophisticated investors. It provides evidence that screening matters. It suggests that when investors are pulled into intense work with existing portfolio companies, the quality of new investment decisions can suffer.

That is the lesson. Attention is not a soft input. Attention is part of underwriting. And if attention is part of underwriting, then investment models that preserve attention may deserve more respect.

The Endurance Capital translation

At Endurance Capital, the same pattern keeps appearing across sport, longevity, founders, and capital. Durability is not intensity. It is pacing. It is knowing when to move and when to wait. It is knowing which signals matter and which signals are only noise. It is resisting the urge to chase someone else’s race.

This is also true in early-stage investing.

The market rewards activity socially. It rewards announcements, logos, speed, and the appearance of access. But the investment return is often determined by something quieter: the quality of attention before the decision.

Did the investor understand the founder? Did they understand the technical risk? Did they understand the financing path? Did they understand the cap table? Did they understand why the opportunity was available? Did they understand what could break? Did they understand what everyone else was missing?

That is the race. Not the announcement. Not the dinner. Not the forwarded deck.

The race is the quality of judgment under uncertainty. And judgment requires attention.

In endurance, pacing is the difference between strength and collapse. In early-stage investing, attention may be the equivalent of pacing.

A practical direct-investment filter

A disciplined direct investor should ask five questions before entering a startup cap table.

First, why is this opportunity available? If the answer is relationship, reputation, founder trust, or clear strategic fit, it may be worth time. If the answer is that the company cannot close the round elsewhere, the work has to become more rigorous, not less.

Second, who wants this investor in the cap table? The founder matters. The board matters. Existing investors matter. Other VCs matter. If strong investors are willing to refer, include, or co-invest with a direct investor, that is a signal. Not proof, but signal.

Third, where is the information asymmetry? The earlier, more technical, more regulated, or more cross-border the company, the more important expert diligence becomes.

Fourth, what does the founder do under challenge? The best founders do not simply perform confidence. They update. They clarify. They defend what matters and adapt what does not.

Fifth, what value is being added beyond capital? Sometimes money is enough. Often, in the best rounds, it is not. The investor needs to know why their presence improves the cap table.

These questions do not remove risk. Nothing removes risk in venture. But they improve the quality of attention. And in early-stage investing, that may be one of the few edges that can still compound.

Closing

Beyond Abuzov and Lerner, the wider academic literature from leading U.S. business schools points in the same direction without overstating the case. Paul Gompers at Harvard Business School, Steven Kaplan at Chicago Booth, and Ilya Strebulaev at Stanford Graduate School of Business show that venture capital is a craft built around sourcing, screening, contracting, monitoring, and post-investment support, not simply the provision of capital. Their survey of 885 institutional VCs across 681 firms is useful because it clarifies what “good venture capital” actually does. Laura Huang at Wharton and Andy Wu at Harvard Business School have studied angel and early-stage investing under uncertainty, including the role of judgment, networks, and qualitative assessment when hard data is limited. Antoinette Schoar at MIT Sloan, together with William Kerr and Josh Lerner, has shown that ventures funded by successful angel groups experienced better outcomes than comparable rejected ventures, including improved survival, exits, employment, patenting, web traffic, and follow-on financing.

The conclusion is not that direct investing is categorically superior to venture funds. The more defensible conclusion is that early-stage outcomes are shaped by attention, screening, networks, judgment, access, alignment, and post-investment value — exactly the variables that determine whether direct cap-table investing is disciplined capital or merely passive exposure.

The venture industry loves to talk about access. Access matters. But access is only the door. What happens next depends on attention: attention to the founder, the terms, the market, the technology, the cap table, what is not being said, and why this opportunity is available now.

Professor Rustam Abuzov’s research gives academic weight to something many investors feel but do not always say clearly: in venture capital, busyness is not neutral. Attention constraints can affect screening. Screening can affect outcomes. And when information asymmetry is high, the cost of distraction rises.

That insight sits directly at the intersection of the direct cap-table investment thesis and the Endurance Capital worldview.

The best capital is not always the busiest capital. The best investor is not always the one who sees the most. Sometimes the better investor is the one who can pay deeper attention to fewer things, earn their place in the cap table, build trust with founders and co-investors, and stay clear when the opportunity is still uncertain.

That is not the easiest way to invest.

But in early-stage venture, easy is rarely where the edge lives.

References

1. Rustam Abuzov, Busy Venture Capitalists and Investment Performance, SSRN working paper.

2. Rustam Abuzov, Busy Venture Capitalists and Investment Performance, Journal of Financial and Quantitative Analysis, Volume 60, 2025.

3. Rustam Abuzov, University of Virginia Darden School of Business faculty profile.

4. Rustam Abuzov, LinkedIn profile.

5. Abuzov’s paper reports that startups funded by VC partners during IPO engagement periods are 9.3% less likely to go public or be acquired and have 18.8% lower exit multiples, with stronger effects in higher workload intensity and higher information asymmetry settings.

6. The supplementary material describes VC involvement in the IPO process, including underwriter selection, board and committee work, drafting IPO prospectus materials, roadshows, IPO pricing, and share allocation.

7. The paper defines screening broadly to include deal sourcing, due diligence, investment analysis, contracting, ability to close, selection criteria, and time with founders.

8. Abuzov’s heterogeneity analysis finds that the effects are stronger when information asymmetry and screening intensity are higher, while the negative effects are not present among follow-on participations or nonlead investments.

Disclaimer

Endurance Capital is intended for educational and informational purposes only. Nothing in this article or accompanying materials should be construed as investment, legal, medical, or professional advice. Views expressed by guests or referenced thinkers are their own.

Editorial note

This article was produced from my original notes, source material, and independent research, and refined with the assistance of AI as an editorial tool. The ideas, conclusions, and final wording reflect my own judgment.

Yes — I would make it more human, less “poster-like,” and easier to read in one sitting.

For Substack performance, the best version is usually not the most compressed one and not the most academic one. It is the one that feels like a smart person speaking directly to the reader, with enough evidence to create trust and enough emotional truth to make people share it.

Here is the rewritten version.

Most difficult conversations fail earlier than we think.

They do not fail when someone raises their voice. They do not fail when the other person says the wrong thing. They do not even fail when there is disagreement.

They often fail before the first sentence is spoken. They fail because nobody has decided what the conversation is actually for.

One person walks in wanting repair. The other walks in wanting proof. One person wants clarity. The other wants escape. One wants accountability. The other wants to win.

And once winning enters the room, the original problem starts to disappear. The conversation becomes a contest. Tone becomes evidence. Memory becomes ammunition. Silence becomes strategy.

Anyone who has been in a hard relationship, a hard company, a hard board meeting, or a hard family conversation knows this feeling. You start with an issue. Somehow, ten minutes later, you are no longer talking about the issue. You are defending your identity.

Who cares more?

Who sacrificed more?

Who was right first?

Who remembers it correctly?

Who has the moral high ground?

At that point, the conversation has stopped being a conversation. It has become a race paced by ego. And this is where endurance has something to teach us. In endurance sport, one of the fastest ways to ruin a race is to forget the objective. You react to someone else’s move. You chase a pace that is not yours. You burn energy to protect pride. You confuse movement with strategy.

Hard conversations work the same way. The moment you forget the ending you want, ego starts pacing the race. So the rule is simple:

Before you try to win, decide what ending you want.

The Roman rule I could not verify

This article started with a viral idea I saw online.

The post claimed that ancient Roman couples had a rule before difficult conversations. Before the argument began, each person was supposedly expected to clarify one thing: what outcome do I actually want from this conversation?

Not revenge. Not dominance. Not proving who was right. Resolution.

It is a beautiful idea. It is exactly the kind of idea that travels well online because it sounds ancient, elegant, and immediately useful. But I could not verify it as an actual Roman marital custom.

I found versions of the claim circulating on social media, but I did not find credible support for it in the classical sources I checked, or in reliable scholarship on Roman domestic life. So I would not present it as historical fact. That matters.

Endurance Capital should not repeat mythology just because the mythology is useful. But sometimes a viral idea points toward something real. And this one does. The exact “Roman rule” may be modern. The deeper principle is ancient.

Roman Stoic writers cared deeply about the problem underneath it: how to keep anger, ego, status, and impulse from taking control of action. Cicero wrote about the difference between settling disputes through discussion and settling them through force. Epictetus began the Enchiridion by separating what is in our control from what is not — our opinions, desires, aversions, and actions. Seneca wrote extensively about anger and the need to interrupt it before it turns into behavior.

So the honest version is this:

The Roman rule may not be real. But the endurance rule is.

Before you speak, decide what ending you want.

Conflict is often a pacing problem

We usually talk about conflict as a communication problem. Sometimes that is true. People use poor words. They interrupt. They accuse. They avoid. They exaggerate. They listen badly.

But underneath the communication problem, there is often a pacing problem. Someone goes out too hard. Someone surges too early or reacts to the other person’s move instead of staying connected to the course, or forgets the plan because the ego got passed.

That is why so many arguments feel strangely familiar. They begin with a problem and become a performance. Two people are no longer trying to solve anything. They are trying not to lose.

In a relationship, this looks like arguing about tone instead of the wound.

In a company, it looks like a founder defending a decision because backing down feels like weakness.

In a boardroom, it looks like everyone protecting their position while the real risk remains unnamed.

In sport, it looks like chasing someone else’s pace because your pride cannot tolerate being passed.

The pattern is the same.

The original objective disappears. The nervous system takes over.

Modern neuroscience gives us some language for this. Research on emotion regulation shows that prefrontal systems are involved in reappraisal, control, and regulation, while the amygdala is involved in emotional salience and threat response. The point is not the overused phrase “amygdala hijack.” The point is more practical: under emotional load, regulation becomes harder, and the quality of the conversation depends on whether you can bring purpose back online.

That is why the question matters.

What ending do I want?

It sounds simple. It is not. Because in the heat of conflict, many of us do not want an ending. We want discharge. We want relief. We want the other person to feel what we feel. We want to restore status. We want to be seen as right.

We want to win. But winning is often a terrible objective for a conversation that needs to preserve something: marriage, friendship, team, company, partnership, cap table, family, future.

The five endings worth choosing

Before a hard conversation, I think there are usually five useful endings to choose from: Repair, Truth, Decision, Boundary and Learning.

The order matters less than the act of choosing. What matters is that you know what kind of conversation you are trying to have before your emotions choose it for you.

If the ending is repair, you do not lead with accusation. You still tell the truth, but the truth is aimed at reconnection, not punishment. You are not trying to win the courtroom. You are trying to protect the relationship.

If the ending is truth, you do not punish honesty. This is especially important inside companies. A team that gets punished for truth will eventually optimize for performance. Everyone will look aligned. Nothing will be real.

If the ending is decision, you do not keep looping emotionally. You name the options, the trade-offs, and the next move. Many teams call something “conflict” when it is really decision avoidance wearing emotional clothing.

If the ending is boundary, you do not negotiate your self-respect. Not every conversation is meant to end in agreement. Sometimes the ending is simply clarity about what you will and will not continue to accept.

If the ending is learning, you do not treat being wrong as defeat. You treat it as adaptation. This may be the hardest one for high performers, because if your identity depends on being right, every correction feels like humiliation.

But durable people are not built by being right all the time.

They are built by recovering quickly when reality corrects them.

Why founders should care

For founders, this is not soft. It is operating discipline.

A founder’s emotional patterns eventually become the company’s operating system. If the founder enters every hard conversation trying to win, the company learns to hide information. If the founder cannot tolerate ambiguity, the company learns to manufacture certainty. If the founder treats disagreement as disloyalty, the company becomes polite and stupid.

And if the founder needs to be the smartest person in every room, the best people eventually stop bringing their full intelligence. This is why conflict quality is a leadership metric.

Not because every conversation should be calm. Not because teams should avoid intensity. The opposite. The best teams can handle intensity because they know what the conversation is for.

They can disagree without turning disagreement into a referendum on belonging. They can surface risk without making the messenger pay. They can move from heat to clarity. That is rare and valuable.

A founder who can do that has an advantage. A CEO who can do that has an advantage. An investor who can do that has an advantage.

Because long-term compounding depends on the quality of repeated conversations under stress.

Why investors should care

Investors like to talk about judgment.

But judgment is not only what you think when everything is calm. Judgment is what remains available when your position is threatened.

A board meeting is often a perfect laboratory for this. The company misses plan. The market moves. A hiring decision fails. A financing risk appears. Suddenly, everyone’s nervous system enters the room.

The conversation can go in two directions.

One direction is performance. The founder minimizes. The investor pressures. The board asks questions that are really accusations. Everyone protects themselves. Nobody quite says the thing that needs to be said.

The other direction is problem-solving. Reality is named. Responsibility is distributed. Options are clarified. The next move becomes visible. The difference is rarely IQ. It is emotional architecture. What is the conversation for?

If the answer is “to assign blame,” people will behave accordingly. If the answer is “to make the next best decision,” the room changes.

This is why the endurance rule matters in capital allocation. A long-term investor should not only ask what they believe about a company. They should ask what kind of conversations that company has when reality gets difficult.

That answer tells you a lot.

Why it matters at home

The same rule applies at home. Especially at home.

The people closest to us are often the people with whom we are least disciplined. We prepare for investor meetings. We prepare for races. We prepare for board presentations. Then we walk into the most important conversations of our lives with no objective, no pacing, no recovery plan, and no idea what ending we want.

And then we are surprised when the same argument returns. A relationship can survive conflict. It cannot survive endless conflict with no intended ending.

Before speaking, ask yourself: do I want to be understood, or do I want to understand? Do I want repair, or do I want an apology? Do I want a boundary, or do I want control? Do I want to stop repeating a pattern, or do I want to win this round?

That question does not make the conversation easy. It makes it honest.

It also reveals something uncomfortable. Sometimes we do not want repair. We want victory. Sometimes we do not want truth. We want confirmation. Sometimes we do not want understanding. We want the other person to finally lose. That is human. But if we do not notice it, it drives the race.

What neuroscience adds

The Stoics did not have fMRI scanners.

But they understood something practical about human beings: emotion can outrun reason, and the work is to create space before action.

Modern research gives us a biological language for part of that process. Studies on affect labeling suggest that putting feelings into words can reduce amygdala activity and engage prefrontal pathways involved in regulation. In plain English, naming what is happening can help move the system from raw reaction toward regulation.

That does not mean one sentence magically fixes conflict. It does not mean every argument can be solved with a communication trick. And it certainly does not mean people should use calm language to avoid accountability.

But it does suggest that the pause matters. The label matters and so does the objective.

“What ending do I want?” is not just a philosophical question. It is a regulation tool. It asks the mind to move from threat to purpose. From impulse to direction. From winning to designing.

The endurance rule

So here is the rule I would keep. Not because it is Roman. Because it is durable.

Before you speak, decide what ending you want.

If you cannot answer, you are probably not ready to have the conversation. Walk. Write. Wait. Breathe. Clarify. Then return.

This is not weakness. It is not avoidance. It is not conflict phobia. It is pacing. And pacing is one of the highest forms of strength.

The strongest people are not the ones who win every argument. They are the ones who can stay clear when the conversation gets hot. They know when ego is trying to hijack the race. They know the difference between intensity and progress. They know that a relationship, a company, a board, or a family can survive conflict. But it cannot survive endless conversations where nobody knows what ending they are trying to reach.

Before you try to win, decide what ending you want. That is not a Roman rule. It is an endurance rule. And in any life that needs to last, it may be one of the most important rules we have.

Five questions to carry into your next hard conversation

What ending do I actually want?

Am I trying to solve the problem, or protect my ego?

What would repair look like here?

What truth needs to be said without punishment?

What would I say differently if I cared more about the relationship than the win?

References

1. The viral “Roman rule” claim appears in social media posts, but I could not verify it as a documented Roman marital custom or classical source. I treat it here as a modern viral framing rather than a historical fact.

2. Cicero, De Officiis, Book I, section 34. Cicero distinguishes between settling disputes by discussion and by force, describing discussion as characteristic of human beings. Source

3. Epictetus, Enchiridion, section 1. Epictetus distinguishes what is in our control from what is not, including opinion, pursuit, desire, aversion, and our own actions. Source

4. Seneca, De Ira / On Anger. Seneca’s treatment of anger focuses heavily on delaying and examining anger before it hardens into action. Source

5. Ochsner, K. N., Silvers, J. A., & Buhle, J. T. “Functional imaging studies of emotion regulation: a synthetic review and evolving model of the cognitive control of emotion.” Annals of the New York Academy of Sciences, 2012. Source

6. Lieberman, M. D. et al. “Putting feelings into words: affect labeling disrupts amygdala activity in response to affective stimuli.” Psychological Science, 2007. Source

7. Burklund, L. J. et al. “The common and distinct neural bases of affect labeling and reappraisal in healthy adults.” Frontiers in Psychology, 2014. Source

Disclaimer

Endurance Capital is intended for educational and informational purposes only. Nothing in this article or accompanying materials should be construed as investment, legal, medical, or professional advice. Views expressed by guests or referenced thinkers are their own.

Editorial note

This article was produced from my original notes, source material, and independent research, and refined with the assistance of AI as an editorial tool. The ideas, conclusions, and final wording reflect my own judgment.

The people who last are rarely the people who do everything alone. They are the people who build the right support system before they need it.

Endurance Capital — Episode 6: Powered by the Pack is now live on YouTube and wherever you listen to your podcasts.

Featuring

Crowie Alexander
3× Ironman World Champion
2× Ironman 70.3 World Champion
Physiotherapist | 25-year professional triathlon career

Anna-Luisa Schaffgotsch
Co-founder & CEO, impli.co

There is a metric in endurance sport that never appears on the watch.

Not watts.
Not pace.
Not heart rate.
Not VO₂ max.

It is quieter than that.

It is the person standing beside you when the race goes badly.
The coach who tells you the truth before you break.
The partner who sees the cost before the public sees the result.
The training group that pulls you back into rhythm.
The co-founder who steadies the room when everyone else is panicking.

In elite performance, we love the individual.

The athlete crossing the line.
The founder on stage.
The CEO making the call.
The investor with conviction when everyone else has left.

But if you look closely at the people who actually last, there is almost always a pack behind them.

That is the theme of Episode 6 of Endurance Capital: Powered by the Pack.

And it may be one of the most important lessons for anyone building something that needs to endure.

The hidden system behind durability

Crowie Alexander knows this better than most.

He is a 3× Ironman World Champion, 2× Ironman 70.3 World Champion, physiotherapist by training, and one of the most durable athletes in the history of triathlon.

But what makes Crowie interesting is not just the winning.

It is the length of the arc.

A 25-year professional career.
A path built from the outside.
No easy federation machine.
No simple junior pipeline.
No straight-line route to the top.

He built a career through family, sponsors, training partners, coaches, discipline, and a support structure designed around continuous improvement rather than one heroic moment.

That matters because most people misunderstand endurance.

They think endurance is the ability to suffer alone.

It is not.

Real endurance is often the ability to build the conditions that make long-term effort survivable.

And those conditions are rarely individual.

The founder version of the pack

In startups, the myth is very similar.

The lone founder.
The heroic operator.
The person who just grinds harder than everyone else.

It makes for good storytelling.

It makes for terrible operating systems.

The founder who has no pack eventually becomes the bottleneck, the pressure point, and the single point of failure.

At first, this can look like strength.

They answer every message.
They own every decision.
They carry every emotional spike.
They absorb every investor rejection.
They keep the company alive through force of will.

But over time, what looks like commitment can become fragility.

Because no one can carry complexity alone forever.

In Episode 6, Anna-Luisa Schaffgotsch, co-founder and CEO of impli.co, brings the founder lens into the conversation.

She is building in women’s health and hormone monitoring — a space that is technical, regulated, deeply human, and long horizon by nature.

That kind of company cannot be built on adrenaline alone.

It needs co-founders, advisors, investors, scientific support, personal support, and a system that keeps the founder emotionally intact enough to keep making good decisions.

The same lesson repeats across sport and startups:

The pack is not a luxury.

It is infrastructure.

Losing well requires witnesses

One of the most important parts of the episode is not about winning.

It is about losing.

Elite athletes lose constantly.

Even the greats.

The difference is not that they avoid loss. It is that they learn how to metabolize it without turning every bad day into an identity crisis.

That is where coaches, family, and support systems matter.

A good pack helps you separate data from identity.

A bad race does not mean you are finished.
A rejected funding round does not mean you are not a founder.
A failed pilot does not mean the company has no future.
A difficult quarter does not mean the thesis is dead.

It means something happened.

Now the work is to extract the signal without collapsing into the story.

The athletes who last are not always the most talented.

Often, they are the ones with the best environment.

The people around them help them recover from losses, avoid emotional overreaction, and return to the work with enough clarity to improve.

Founders need the same thing.

So do investors.

So do chief executives.

Because in any long game, the loss itself is not always what breaks people.

It is the meaning they attach to the loss.

The pack helps correct that meaning.

Support by design, not accident

One of the most practical ideas in this episode is simple:

Do not wait until you are breaking to build support.

By then, it is usually too late.

Athletes understand this better than founders.

A serious athlete does not wait until they are injured to find a coach.
They do not wait until race week to think about nutrition.
They do not wait until burnout to plan recovery.
They build support into the program.

Founders often do the opposite.

They wait until the company is on fire.
Until the relationship is strained.
Until sleep is broken.
Until decisions feel foggy.
Until they cannot tell the difference between urgency and panic.

Then they look for support.

But support works best when it is designed early.

That means asking honest questions before the crisis:

Who tells me the truth when I am overreaching?
Who helps me recover when I lose?
Who sees the signals before I admit them?
Who understands the cost of what I am building?
Who can challenge me without threatening me?
Who helps me stay in the game?

The pack is not just a group of people who cheer for you.

That is encouragement.

The pack is a system that helps you remain durable.

The wrong pack can break you faster

There is another side to this.

Not all support is supportive.

Some people amplify panic.
Some normalize overwork.
Some reward self-destruction.
Some confuse intensity with seriousness.
Some want proximity to the win but disappear in the difficult middle.

The right pack helps you become more stable.

The wrong pack makes your nervous system more chaotic.

This matters in sport.

It matters even more in startups.

Because the founder’s environment becomes the company’s environment.

If the people around the founder reward constant panic, the company becomes frantic.

If investors only reward speed without durability, the founder starts confusing motion with progress.

If the leadership team avoids honest conversations, the company loses the ability to adapt.

A good pack is not soft.

It is not a comfort blanket.

It is a performance system.

It helps you take the right risks, recover from the wrong ones, and stay honest when the stakes rise.

What founders can learn from Crowie

Crowie’s story is powerful because it reminds us that long careers are not built only from big wins.

They are built from thousands of decisions that make the next season possible.

Who you travel with.
Where you train.
How you recover.
Who you trust.
What you do after bad races.
How you keep your identity from becoming too fragile.
How you protect the relationships that protect you.

Founders and chief executives should pay attention to that.

Because company-building is also a long race.

And the brutal middle is where many people quietly disappear.

Not because they lacked talent.

But because they never built the system that allowed the talent to last.

The lesson for builders and investors

For founders, the lesson is direct:

Build your pack before you need rescuing.

That means co-founders who can carry truth, not just tasks.
Investors who understand pacing, not only pressure.
Partners and friends who are allowed to see the cost.
Advisors who can separate a bad week from a bad strategy.
A team culture where support is not seen as weakness, but as design.

For investors, the lesson is just as important:

Do not only assess the founder.

Assess the founder’s system.

Who is around them?
Who helps them think clearly?
Who challenges them?
Who protects them from their own worst instincts?
Who will still be there after the first serious loss?

The quality of the pack often predicts the durability of the builder.

And durability is one of the rarest assets in a long game.

Closing reflection

We like to think endurance is solitary.

The lonely road.
The silent swim.
The final miles where nobody can do the work for you.

And that is partly true.

At some point, every athlete has to run their own marathon.
Every founder has to make their own decision.
Every leader has to carry the responsibility that cannot be delegated.

But nobody reaches that point alone.

The people who last are not carried the whole way.

They are held together long enough to keep choosing the work.

That is the pack.

And in sport, startups, investing, and life, it may be the quiet difference between a brilliant season and a durable career.

Sponsor

This episode is brought to you by Stride.

One thing I’ve learned from athletes, coaches, founders, and everyone in between is that understanding your own biology is the real unlock.

Stride helps you do exactly that.

It brings your sessions, your data, and your goals into one place, so you can see what is working — and what is not.

Use code TRAMPOLINE for:

£100 off StrideOne
30% off any other individual Stride tests or supplements

Visit GetStride.com.

Disclaimer

Endurance Capital is intended for educational and informational purposes only. Nothing in this episode or accompanying materials should be construed as investment, legal, medical, or professional advice. Views expressed by guests are their own.

Editorial note

This article was produced from original interviews and source material and refined with the assistance of AI as an editorial tool. The ideas, conclusions, and final wording reflect my own judgment.

AI may not just change how we work. It may change what kind of humans we become.

There’s a question sitting quietly underneath the AI boom that almost nobody seems willing to ask.

Not: Will AI replace jobs?
Not: Will AI make us more productive?
Not even: Will AI become more intelligent than us?

The deeper question is this:

What happens to human beings when friction disappears from cognitive life?

Because beneath all the excitement around artificial intelligence lies something much more biological.

Something about endurance.
Something about adaptation.
Something about what happens to the nervous system when struggle becomes optional.

And if there’s one thing Endurance Capital has explored from Pacing the Impossible to the upcoming Powered by the Pack, it is this: durable people are not built by removing all stress. They are built by learning how to metabolize it. The series has consistently translated endurance, adaptation, pain, longevity, and support systems into practical frameworks for founders, investors, and high performers.

AI as the new training environment

Kashmir Hill, the New York Times technology reporter and author of Your Face Belongs to Us, has spent years examining the unintended consequences of technology, privacy, facial recognition, and now AI. Her work focuses on the unexpected and sometimes ominous ways technology changes human life. Her personal website describes that work here. (Kashmir Hill)

That lens matters here.

Because AI is not simply another productivity tool.

It is the first mainstream technology that competes directly with human cognitive effort.

It can write before we struggle to formulate.
Summarize before we sit with complexity.
Answer before we remember.
Reassure before we tolerate uncertainty.

And effort changes us.

The biology of effort

Endurance athletes understand something Silicon Valley often forgets:

Stress is not the enemy.

Unrecovered stress is.

The entire logic of training is based on calibrated discomfort. You expose the body to strain, recover, and adapt. Too little stress and nothing changes. Too much stress and the system breaks.

The same is true for the mind.

Attention, memory, judgment, emotional regulation, and self-trust are not static traits. They are trained through use.

This is where Dr. Alia Crum becomes important. Crum is a Stanford psychology professor and principal investigator of the Stanford Mind & Body Lab, where her research explores how mindsets shape behavior, physiology, health, and performance. Stanford describes her work as examining how mindsets can alter objective reality through behavioral, psychological, and physiological mechanisms. (Alia Crum, PhD)

One of Crum’s most useful ideas for this conversation is that stress is not only harmful by default. How we perceive and engage with stress can change its effect on us. Her work is highly relevant to any discussion about whether removing too much difficulty from human life makes us safer — or simply less adapted. (Stanford Psychology)

That is the bridge to AI.

If every difficult cognitive moment is outsourced, we may not simply become more efficient.

We may become less adapted.

The real risk: not intelligence loss, but durability loss

The danger of AI may not be that it makes us less intelligent.

It may be that it makes us less durable.

A founder who never sits alone with uncertainty may lose strategic depth.
An investor who outsources synthesis may weaken pattern recognition.
A student who never wrestles with a blank page may lose the ability to think in long arcs.
A human who turns to AI for every moment of discomfort may slowly lose tolerance for friction.

This is not anti-AI.

It is biology.

The nervous system adapts to what it repeatedly experiences.

If AI becomes an infinite layer of cognitive outsourcing, humans may slowly lose deep focus, memory endurance, reflective judgment, boredom tolerance, emotional regulation, and the ability to metabolize uncertainty independently.

The issue is not whether AI is useful.

It is whether we use it in a way that preserves human adaptive capacity.

Cognitive Zone 2

Endurance athletes understand Zone 2.

It is the steady aerobic pace where athletes build endurance without overreaching.

Low enough intensity to sustain.
Hard enough to build the engine.

Maybe the AI era needs the same idea.

Call it Cognitive Zone 2.

Not rejecting AI.
Not worshipping AI.

Using it while deliberately preserving the human work that keeps us strong.

That might mean:

Writing before prompting.
Thinking before summarizing.
Reading deeply before asking for extraction.
Letting boredom exist.
Memorizing some things.
Having hard conversations without mediation.
Doing difficult work before asking for acceleration.
Training the body to regulate the mind.

Because the biology of effort matters.

Always has.

The founder and investor translation

For founders, AI will be extraordinary leverage.

But leverage without durability is dangerous.

The strongest operators in the AI era may not be the ones who automate everything first. They may be the ones who understand what is worth automating — and what must remain deeply human.

Judgment.
Taste.
Conviction.
Pacing.
Self-trust.
Emotional regulation.
Moral responsibility.
The ability to suffer intelligently without collapsing.

Those are not inefficiencies.

They are the operating system.

The philosophy behind Endurance Capital

Endurance Capital has never really been about sport alone.

It is about what elite performance teaches us about building and backing durable things.

Pacing the Impossible was about restraint, conviction, and not spending the whole race in the first mile.
Fluid Intelligence was about staying clear when the plan breaks.
Pain to Performance was about learning to read discomfort as information.
The Engine Room for Endurance was about the biological systems that allow us to go long.
Built to Last, Not to Break was about durability rather than self-destruction.
And Powered by the Pack, still to come, is about the support systems that keep people in the game for decades.

AI now sits inside all of those questions.

How do we pace ourselves when machines accelerate everything around us?
How do we preserve fluid intelligence when answers arrive before reflection?
How do we distinguish useful relief from the removal of developmental stress?
How do we build bodies, minds, companies, and lives that last in an age designed to make everything faster?

This is why AI is not only a technology conversation.

It is also a human-performance conversation.
A longevity conversation.
An endurance conversation.

Closing

The future may belong to people who can use AI without becoming dependent on it.

People who can move faster without losing depth.
Automate without atrophying.
Delegate without disappearing.
Think with machines without forgetting how to think alone.

Because eventually, the real advantage may not be artificial intelligence.

It may be human resilience.

And that may be the longevity question nobody is asking about AI.

Further reading

• Kashmir Hill — personal website

• Kashmir Hill — bio

• Alia Crum — personal website

• Alia Crum — Stanford Psychology profile

• Stanford Mind & Body Lab

Disclaimer

Endurance Capital is intended for educational and informational purposes only. Nothing in this episode or accompanying materials should be construed as investment, legal, medical, or professional advice. Views expressed by guests are their own.

Editorial note

This article was produced from original interviews and source material and refined with the assistance of AI as an editorial tool. The ideas, conclusions, and final wording reflect my own judgment.

If Episode 4 was about longevity, Episode 5 is about something more practical:

how to build a career that actually lasts — not just one that peaks.

That is what this Endurance Capital conversation with Daniela Ryf made clear.

Follow Episode 5 | Built to Last, Not to Break here:
▶️ YouTube → Episode 5 | Built to Last, Not to Break
🎧 Listen on Spotify, Apple Podcasts, Amazon Music and Deezer

When people think about elite performance, they usually picture the obvious things:

winning, talent, intensity, domination.

But that is not what stood out most in Daniela.

What stood out was durability.

Not one great race.
Not one perfect season.
Not one peak people remember forever.

A more difficult form of excellence:

holding a high standard for long enough that it becomes a career, a body of work, and eventually a reputation.

Daniela Ryf is one of the defining long-course athletes of her era: a five-time IRONMAN World Champion and five-time IRONMAN 70.3 World Champion. That matters not just because of the titles, but because of what they represent: repeated excellence under repeated pressure.

That is what makes this conversation relevant far beyond triathlon.

In startups, the equivalent is obvious.

Anyone can look strong in the first sprint.
Anyone can look serious in the launch phase.
Anyone can look dangerous when the market is still surprised by them.

The harder question comes later.

Can you keep producing when the standard is already high?
Can you keep evolving when everyone is chasing the version of you that already won?
Can you hold the pace without hollowing yourself out in the process?

That is where Daniela’s story becomes useful.

Because a lot of ambitious people know how to surge.
Far fewer know how to sustain.

The lesson is not simply:

work harder.

It is something more demanding:

build in a way that survives your own ambition.

That is the difference.

A company can be built on urgency for a while.
A team can be carried by force of will for a while.
A reputation can rise quickly on one period of extraordinary output.

But “for a while” is not the same thing as lasting.

And that is the distinction this episode sharpens.

The real challenge is not producing one extraordinary result.

It is building something structurally sound enough to keep producing under pressure.

That is true in triathlon.
It is true in business.
And it is true in life.

The people who last usually learn the same lesson:

what matters is not only how high you can go, but what you have to protect in order to keep going.

Energy.
Health.
Focus.
Identity.
Relationships.
The ability to adapt without lowering the bar.

That is what “built to last, not to break” really means.

Not success as a moment.
Success as a structure.

Three practical takeaways

1. Do not confuse intensity with durability.
A lot of people can push hard for a short period. Far fewer can hold a high standard for years without becoming brittle.

2. What gets protected is what keeps performing.
Energy, health, recovery, identity, and focus are not soft considerations. They are part of the system.

3. Winning creates a second challenge.
The task is not just getting to the top. It is staying functional once you are there.

The philosophy behind Endurance Capital

This is the core idea behind Endurance Capital:

the lessons that shape durable champions are often the same ones that shape durable founders and investors.

Not because sport and business are identical.

But because both reward people who can manage pressure, fatigue, timing, sacrifice, and decision-making over long periods of time.

Endurance Capital sits at that intersection.

It is not about motivation.

It is about applied thinking under pressure.

Closing

This episode is really about one question:

What does it take to stay at the top long enough to build something that lasts — not break under the weight of it?

That is where serious performers separate from reactive ones.

And where durable founders separate from people who confuse intensity with strength.

Episode 5 of Endurance Capital is built around that distinction.

With Daniela Ryf, it becomes a conversation about standards, sacrifice, and the harder form of excellence: staying structurally sound enough to keep producing when the pressure no longer feels exciting, only familiar.

If you build, invest, or lead through long cycles, start here.

Disclaimer

Endurance Capital is intended for educational and informational purposes only. Nothing in this episode or accompanying materials should be construed as investment, legal, medical, or professional advice. Views expressed by guests are their own.

Editorial note

This article was produced from original interviews and source material and refined with the assistance of AI as an editorial tool. The ideas, conclusions, and final wording reflect my own judgment.

There are regimes where capital rewards scale.

There are regimes where capital rewards liquidity.

And there are regimes where capital begins to reward adaptation speed.

We are entering the third regime.

The UAE’s announcement that 50% of government sectors, services, and operations will run on agentic AI within two years is easy to misread as another technology initiative.

It is not.

It is a state-capacity signal.

Under the directives of President Sheikh Mohamed bin Zayed Al Nahyan, the UAE is moving toward a government model where autonomous systems do not simply assist existing processes. They analyse, decide, execute, and improve in real time.

This is not digitisation.

This is not e-government.

This is the beginning of an AI-native administrative state.

The important shift is not that government will use AI. The important shift is that government work is being redesigned around AI.

Most countries are still debating adoption frameworks.

The UAE has set a deadline.

Two years.
Fifty percent of government operations.
Named senior leadership.
A dedicated execution taskforce.
Performance measured by speed, quality, and mastery of AI implementation.

That is not a press release.

That is an operating cadence.

The new sovereign advantage

For most of the post-war period, investors evaluated jurisdictions through familiar variables: rule of law, currency stability, taxation, labour depth, education, infrastructure, and market access.

Those variables still matter.

But a new one is rising:

decision velocity.

A state that reduces administrative friction lowers the time cost of capital.

A state that makes decisions faster increases the velocity of enterprise formation.

A state that coordinates policy, infrastructure, and talent at speed becomes more attractive to founders, family capital, strategic industries, and sovereign partnerships.

This is why the UAE announcement matters.

The country is not simply trying to make government more efficient.

It is trying to turn execution speed into a national asset.

The macro backdrop

The UAE is making this move from a position of strength.

The Central Bank of the UAE projected real GDP growth of 5.6% for both 2025 and 2026, supported by non-hydrocarbon sectors, financial services, manufacturing, construction, and hydrocarbon recovery.

That compares with much lower growth across many advanced economies. In the same period, global growth is expected to remain near the low-3% range, while several mature economies are projected to grow closer to 1–2%.

The more important point is not the growth differential itself.

It is the composition.

The UAE Ministry of Economy reported that non-oil activities reached 77.3% of real GDP in Q1 2025, with non-oil GDP growing 5.3% year-on-year.

That matters.

The UAE is not using AI to compensate for a weak economic model.

It is adding AI to an already diversified execution machine.

Logistics.
Finance.
Tourism.
Real estate.
Manufacturing.
Technology.
Capital mobility.
Talent inflows.

Agentic government is the next layer.

From digital services to autonomous execution

States do not become AI-native by announcement.

They need infrastructure beneath the ambition.

Digital identity.
Clean records.
Data-sharing systems.
Interoperable agencies.
Service redesign.
Trained employees.
Clear accountability.
Leadership authority.

The UAE has spent years building this administrative substrate through e-government, mobile government, UAE Pass, proactive services, digital records, and national AI strategy.

The new agentic AI framework sits on top of that foundation.

This is why the announcement should be taken seriously.

Without infrastructure, AI in government becomes theatre.

With infrastructure, it becomes execution capacity.

The overlooked detail: training the state itself

The most important part of the UAE announcement may not be the 50% target.

It may be the commitment to train every federal employee to master AI.

That line matters because AI productivity does not come from software alone.

It comes from institutional absorption.

The employee has to understand the tool.
The manager has to redesign the workflow.
The ministry has to measure the result.
The system has to reward adoption rather than preserve inertia.

This is the difference between buying AI and becoming AI-native.

The UAE is not merely acquiring technology.

It is attempting to retrain the administrative class around a new operating system.

What markets may be mispricing

Markets still treat government quality as mostly static.

It is not.

State capacity can compound.

A jurisdiction that improves its administrative operating system every year becomes meaningfully different over a decade.

It becomes faster to start companies.
Faster to move talent.
Faster to approve projects.
Faster to regulate emerging industries.
Faster to allocate capital.
Faster to respond when the terrain changes.

The second mispricing is assuming that larger legacy systems will automatically dominate the AI transition because they have deeper universities, larger capital markets, and bigger populations.

Those advantages remain real.

But AI rewards coordination as much as scale.

A smaller state with aligned leadership, fiscal capacity, digital infrastructure, and high tolerance for administrative redesign may move faster than larger systems burdened by institutional self-protection.

The third mispricing is moral vocabulary replacing strategic analysis.

The useful question for capital is not ideological.

It is operational:

Which systems can reduce friction?
Which systems can retrain their people?
Which systems can deploy infrastructure?
Which systems can make decisions before the window closes?

Capital is not sentimental.

It notices where time is being saved.

The Endurance Capital lens

In endurance, the decisive athlete is rarely the one with the highest peak output.

It is the one who can absorb stress, recover quickly, and change pace without structural breakdown.

The same is true for capital systems.

A state that can adapt under pressure has durability.

A state that requires too much internal negotiation before every change may still look stable, but its stability can harden into rigidity.

Endurance Capital is not speed for its own sake.

It is the capacity to preserve optionality while moving fast enough for the terrain.

That is what the UAE is testing.

Not whether AI can answer questions.

Whether AI can increase the cadence of the state.

The uncomfortable implication

The AI transition may expose a weakness that previous industrial transitions did not.

Industrial growth rewarded infrastructure, labour, capital depth, energy, education, and property rights.

Large advanced systems were strong across many of those variables.

AI rewards something slightly different:

coordinated execution speed.

It is not enough to have excellent universities.

It is not enough to have venture capital.

It is not enough to publish policy papers.

The state has to metabolize the technology.

And the UAE is attempting to do that at national scale.

AI will not only separate companies by productivity. It will separate states by execution capacity.

This does not mean fast systems always win.

Speed creates risks.
Centralized execution can miss feedback.
Autonomous systems can fail in ways that are difficult to audit.
Public trust matters.
Human judgment matters.
Accountability matters.

But the strategic asymmetry is clear.

In a world where many systems are becoming heavier, slower, and more internally constrained, the ability to move is becoming a form of power.

Capital allocation implications

This is not a recommendation to allocate blindly to one geography.

It is a framework for evaluating sovereign operating systems.

For family offices, founders, and cross-border investors, the question is changing from:

Where is capital protected?

to:

Where can capital still move?

The old hierarchy of jurisdictions was built around institutional predictability.

The new hierarchy will also include administrative velocity.

That affects where founders incorporate, where families establish operating presence, where venture platforms source deal flow, where regulated products are tested, and where strategic capital builds partnerships.

The UAE is positioning itself as a jurisdiction where government is not merely a regulator of innovation.

It is an early adopter of its operating logic.

That is a powerful signal.

What would change my mind

Four developments would weaken this thesis.

First, if agentic AI produces visible service failures, cyber vulnerabilities, bias scandals, or accountability gaps that reduce public trust.

Second, if larger advanced systems find credible ways to deploy AI across government without procedural paralysis.

Third, if AI improves interfaces but fails to reduce real approval times, service backlogs, operating costs, or public-sector workload.

Fourth, if geopolitical instability overwhelms domestic execution gains and raises the risk premium attached to the region.

These are the right risks to monitor.

The announcement is ambitious.

Implementation is the test.

Forward signal

The next two years will reveal whether AI-native government is a slogan or a structural advantage.

The key indicators will not be speeches.

They will be processing times, approval cycles, service backlogs, employee productivity, citizen satisfaction, and cross-ministry data interoperability.

If the UAE can show measurable improvement across those dimensions, the implications will travel far beyond public administration.

It will show that execution velocity has become a sovereign advantage.

And in the next phase of global capital allocation, the winners may not be the systems that looked strongest at the start.

They may be the ones that can still move.

If Episode 3 was about turning pain into performance, Episode 4 zooms out to a longer horizon: not just how to endure a hard season, but how to build a system that lasts for decades. With Natalie Coles and Will Harborne, this conversation explores what supercentenarians reveal about stress, rhythm, community, biology, and the future of lifespan.

Follow Episode 4 | The engine room for endurance on Youtube 🎥.

🎧 Listen on Spotify, Apple Podcasts and Amazon Music.

If Episode 3 was about how high performers handle pain, Episode 4 asks a quieter and more ambitious question:

How do you stay in the game for longer than most people think is possible?

That is what this Endurance Capital conversation with Natalie Coles and Will Harborne is really about. Natalie has spent years studying supercentenarians — people who live past 110. Will approaches the same territory from the investor side, asking what extreme human longevity can teach us about biology, intervention, and the future of lifespan science.

What makes the episode useful is that it does not treat longevity as wellness theatre or sci-fi fantasy.

It treats it as a long-game discipline.

Not just living longer in the abstract.
But staying sharper, more functional, and more biologically resilient for longer than the default curve suggests.

That makes Episode 4 a natural continuation of the series.

Episode 2 was about adapting when the plan breaks.
Episode 3 was about reading pain correctly when the cost becomes real.
Episode 4 asks the next long-horizon question:

What if endurance is not only mental, but biological?

That shift matters.

Because many people still think about longevity in the wrong order. They jump straight to future therapies, expensive tools, and exotic interventions. But what stood out most in Natalie’s account of supercentenarians was something quieter:

less drama
more rhythm
more connection
a lighter grip on stress

Again and again, the pattern is striking. These are people who seem to see the storm without living inside it. They are often surrounded by family, embedded in routines, and supported by systems of trust that reduce chronic strain rather than amplify it.

If you treat your nervous system like a war zone for decades, your odds of aging well are probably not great.

That idea runs underneath much of the conversation.

And it lands well beyond health.

What Natalie describes in supercentenarians is not merely old age. It is preserved function. Plenty of people live longer. Far fewer live longer while staying mentally present, biologically stable, and relatively free from the usual cascade of age-related decline.

That is also why María Brañas is such a compelling case in the episode.

Not because she is a miracle.
And not because everyone should copy one food item and expect extraordinary results.

But because her life and data make one thing visible:

it is possible for a human to reach extreme age without the full conventional pattern of breakdown.

The yogurt detail is interesting, of course.

But the deeper lesson is not “eat this and live to 117.”

It is that the microbiome, inflammation, rhythm, diet, and environmental stability may matter more than most people realise — especially when compounded over decades.

That is where Will Harborne’s perspective becomes especially useful.

He does not romanticise lifestyle alone. Two people can live similarly and still age very differently. Biology matters. Genetics matter. And the future of lifespan will likely include interventions far beyond sleep, walking, and yogurt.

But he also makes something else clear:

before the advanced therapies arrive, there is still a great deal available now.

Exercise.
Sleep.
Stress regulation.
Metabolic health.
Immune health.
Mitochondrial function.
Basic consistency.

That is an important correction.

Because longevity can easily become a fantasy category, where people skip over the boring essentials and jump straight to future treatments.

This episode argues for the opposite sequence:

Get the basics boringly right first.
Then, when better tools arrive, you may actually be in good enough shape to benefit from them.

That is a very Endurance Capital idea.

In investing, in sport, and in biology, the glamorous answer is usually overvalued. The compounding answer is usually quieter.

And that is why one of the strongest ideas in this episode is not technological at all.

It is baseline management.

Natalie keeps coming back to consistency.
Will keeps coming back to mechanisms.
Ignacio keeps translating both back to the audience:

know your engine
measure what matters
intervene earlier
play the long game on purpose

That makes the episode especially useful for founders.

Because the founder version of longevity is not necessarily reaching 110.

It is building a body, mind, and operating rhythm that still works in your 50s, 60s, and 70s — instead of winning one decade and paying for it in the next.

The goal is not just to last longer. It is to stay useful for longer.

That is the more practical ambition.

And it applies just as much to investors.

Great investors often talk about duration, patience, and compounding in capital. This episode quietly reminds us that there is a human version of that too.

Your judgment compounds.
Your biology compounds.
Your habits compound.
Your relationship to stress compounds.

So does neglect.

And that may be the clearest lesson from the whole conversation:

Longevity is not just about lifespan. It is about whether your internal systems stay investable over time.

That is why community matters so much in Natalie’s description of supercentenarians. They are rarely alone. They are often protected. They have support. They remain embedded in human systems that lower stress rather than intensify it.

This matters more than ambitious people often admit.

Because isolation is expensive.
And chronic stress is rarely free.

What high performers often call focus is sometimes just unexamined strain.

This episode offers a better standard:

a calmer nervous system
a cleaner rhythm
a more deliberate relationship with data
a stronger support group
and enough biological awareness to see decline before it becomes identity

That last point may be the most practical of all.

One of the best parts of the conversation is the emphasis on measuring your own baseline.

Not obsessively.
Not neurotically.
But longitudinally.

If you can understand your biomarkers over time, rather than waiting for something to go visibly wrong, you give yourself a chance to notice trend before collapse. That is as true in biology as it is in training or business.

Which leads to the deepest idea in the episode:

the most important thing you may leave your children is not only capital.

It may be data.

Because that reframes longevity from private optimisation into intergenerational leverage. Knowing more about your own biology may not only help you. It may help the people who come after you navigate their own risks earlier and more intelligently.

That is where the episode widens beyond health.

It becomes a conversation about stewardship.

Of the body.
Of time.
Of future optionality.

Will Harborne brings in the frontier from there: gene therapies, stem-cell therapies, and interventions that may eventually change what healthy aging looks like at scale. Natalie brings the grounding: supercentenarians are real proof that the human organism can go much further than average aging suggests, but we should not overclaim what lifestyle alone can do.

That tension is healthy.

It keeps the episode honest.

There is hope here, but not hype.
There is ambition here, but not fantasy.
There is practical advice, but no false promise.

Get the basics boringly right for a long time. Then be ready for the next wave of lifespan technology when it arrives.

That may be the practical takeaway.

For athletes: train with biology in mind, not just output.
For founders: stop treating stress as a neutral input.
For investors: think about human durability the way you think about capital durability.
For everyone else: treat rhythm, community, and biological awareness as part of the real work, not as extras.

Episode 4 of Endurance Capital is not really about immortality.

It is about something more grounded and more useful:

how to become a better long-term organism.

And in a world obsessed with intensity, that might be one of the more valuable edges left.

Disclaimer
Endurance Capital is intended for educational and informational purposes only. Nothing in this article, episode, or accompanying materials should be construed as investment, legal, medical, or professional advice. Views expressed by guests are their own. Medical and scientific themes discussed here are exploratory and educational and should not replace individualized professional guidance.

Editorial note
This article was produced from the original Episode 4 conversation and source material and shaped in the establishedEndurance Capital format. It follows the same structure and editorial logic as the previous Episode 2 and Episode 3 follow-up articles: title, subtitle, deck, opening thesis, article body, pull quotes, and closing disclaimers. It was refined with the assistance of AI as an editorial tool; the ideas, framing, and final wording reflect my own judgment.

The lazy interpretation of this market is that investors are hiding in the same giant technology names again.

The better interpretation is that they are doing something far more consequential.

Even in a world shaped by war, tariffs, energy risk, and political fragmentation, capital keeps flowing toward technology. Not because investors have become carefree, but because technology is no longer being priced as a sector in the old sense. It is being priced as infrastructure. As productive capacity. As geopolitical leverage.

That distinction matters.

When a market keeps rewarding technology through conflict, inflation anxiety, and strategic fracture, it is usually telling you that the asset class has become more central to the system, not less. Compute, semiconductors, software, automation, and digital networks are no longer optional growth categories. They are becoming the industrial base of the next decade.

The headlines are about disorder. The market is allocating toward control, productivity, and strategic capability.

The consensus narrative is now too small

Most commentary is still stuck in first-order thinking.

Nvidia is up. Microsoft is strong. Amazon, Alphabet, and Apple remain dominant. Therefore the story must still be narrow leadership, concentrated enthusiasm, and AI euphoria.

That is incomplete.

Yes, mega-cap technology still leads. But the more important question is no longer whether the giants remain in control.

It is whether the market is starting to reward the wider economic system around them.

That is where this gets interesting.

Mega-cap leadership still matters because the core engine is still running

The first thing to say clearly is that the leadership at the top is real.

Microsoft, Nvidia, Amazon, Alphabet, and Apple still sit at the most valuable control points in the stack: compute demand, cloud distribution, enterprise integration, model deployment, and global platform reach. These are not simply crowded momentum trades. They are the best-capitalized entities in the world competing for the infrastructure layer of the next economic cycle.

The semiconductor chain supports the same conclusion. The signal from ASML and TSMC is hard to miss: this cycle is still being funded, and advanced manufacturing capacity remains scarce and strategic.

That matters because broad rallies do not sustain unless the core engine is credible.

And right now, it is.

Finviz S&P 500 one-month heatmap

The S&P 500 heatmap makes the point visually. Mega-cap technology still dominates the map. Semiconductors remain one of the primary leadership groups. Software infrastructure and selected hardware names are clearly participating. This is still a market that pays a premium for scale, cash flow, technical bottlenecks, and ownership of critical digital infrastructure.

But that is no longer the whole story.

The giants are still leading, but leadership at the top is now enabling leadership beneath them.

The more important shift is that the market is beginning to broaden

A narrow rally can be dismissed as concentration.

A broadening rally is harder to dismiss.

What the market appears to be rewarding now is not only the first-order winners of AI, but the wider architecture around them. Not just the chip designers and hyperscalers, but the tools, suppliers, software layers, industrial systems, logistics platforms, and financial rails that become more valuable as the buildout deepens.

This is the subtle shift beneath the headlines.

The Russell 2000 heatmap is useful here, not as a formal dataset, but as directional market texture. Participation is visible across smaller technology names, industrials, regional financials, specialized healthcare, and logistics-adjacent businesses. That does not mean everything is working. It clearly is not. But it does suggest that capital is moving outward from the most obvious winners into the companies that help the system function.

That is a very different signal from a purely narrow AI trade.

Finviz Russell 2000 one-month heatmap
This is why “AI infrastructure” is now too narrow a phrase if it only means GPUs and data centers.

AI infrastructure includes semiconductor tools, networking, power management, cybersecurity, industrial software, warehouse automation, freight optimization, and payments infrastructure. Once you see the cycle this way, the market action becomes more coherent. The market is not simply rewarding intelligence at the top of the stack. It is beginning to reward throughput across the system.

That broadening is the signal.

This is starting to look less like a speculative spike and more like an ecosystem buildout.

Second-order beneficiaries matter more than another Nvidia headline

The next phase of this cycle will be defined by the connective tissue.

Every major technological buildout starts with obvious bottlenecks. The first money goes to the scarce assets everyone can see. The more durable opportunities often emerge one layer deeper, where complexity rises, competition falls, and dependence increases.

That is where second-order beneficiaries come in.

Software that helps enterprises absorb AI into workflows. Industrial automation that turns intelligence into labor productivity. Logistics systems that lower friction across supply chains. Cybersecurity that protects more distributed infrastructure. Digital payments and financial rails that support increasingly automated transactions. Specialized suppliers that sit inside manufacturing and deployment constraints.

Those are not side stories.

They are the operating system of adoption.

And the market appears to be slowly acknowledging that.

This also explains why the rally does not feel indiscriminate. Some defensive areas still lag. Parts of healthcare remain mixed. Parts of energy are uneven. Consumer staples are hardly signaling enthusiasm. The market is not buying everything. It is discriminating toward the places where productivity, automation, and digital leverage are improving the economic equation.

That is usually what a serious rally looks like.

What this means for investors and founders

For investors, the implication is not to chase every company that mentions AI.

It is to understand where value accrues as adoption spreads.

The right question now is not who gets the headline. It is who controls a bottleneck, owns a workflow, reduces cost, compresses time, or becomes strategically hard to replace. In the next phase of this cycle, those attributes will matter more than narrative proximity to Nvidia.

For founders, the lesson is equally sharp.

Being “an AI company” is not the point. Building something that makes the AI economy usable is the point. Workflow integration, compliance, payments, security, logistics, industrial automation, and domain-specific software may prove more durable than generic application stories.

That is where substance begins to outrun fashion.

Finviz whole market sector performance heatmap.

The full-market sector heatmap reinforces this. Leadership is not random. It clusters around artificial intelligence, cloud computing, semiconductors, industrial automation, robotics, logistics, fintech, and selected digital platforms. That pattern matters because it suggests the market is beginning to price a broader industrial and software buildout, not just a handful of celebrity stocks.

When the market stops rewarding only the obvious winners and starts rewarding the ecosystem, it is usually signaling that the buildout is real.

The real signal is deeper than the headlines

In a fractured world, technological power is becoming economic power and geopolitical power at the same time.

That is why war, tariffs, and energy shocks have not displaced technology from the center of markets. If anything, they have reinforced its strategic value.

So the question is no longer whether the giants are still strong.

They are.

The more important question is whether leadership continues to spread through the architecture around them. If it does, the market will be telling us something bigger than “tech is back.” It will be telling us that this cycle is broader, earlier, and more economically embedded than much of the mainstream narrative still understands.

That is what serious investors should be studying now.

Because when the market stops rewarding only the obvious winners and starts rewarding the ecosystem, it is usually signaling that the buildout is real.

Endurance Capital Episode 3 has introduced the theme of pain. Today I write about a deeper follow-up. If Episode 2 gave us a framework for adapting to external change, Episode 3 gives us a framework for adapting to internal strain. The best performers do not simply tolerate pain better. They interpret it better.

If Episode 2 of Endurance Capital was about fluid intelligence — adapting when the middle stops making sense — then Episode 3 | Pain with purpose asks the natural next question:

What do you do when the middle starts to hurt?

That is why this conversation matters.

Pain with purpose | Pain to Performance is a conversation about reframing pain rather than glamorizing it. This piece goes one level deeper. Because the real lesson from this episode is not simply that elite performers can take more pain.

It is that they read it better.

And in this episode, that lesson comes from people with real standing in their worlds.

Mirinda Carfrae is one of the greatest female Ironman champions of all time: a three-time IRONMAN World Champion, IRONMAN 70.3 World Champion, and one of the fiercest runners the sport has ever seen. IRONMAN’s own history notes that in 2013 she won Kona with a faster run split than the men’s champion, Frederik Van Lierde, while setting a new run course record. IRONMAN has also described her as one of the best runners in the sport “in both the women’s and men’s field.”

“The best performers do not panic when pain arrives. They narrow the frame, read the signal, and decide what the next good action is.”

Ian O’Brien has coached elite international squads with a focus on the Olympic Games and World Triathlon Championship Series. World Triathlon’s coach profile highlights his work with top-level athletes, and USA Triathlon named him 2021 Olympic Coach of the Year. Origin Performance describes him as an Olympic medalist, multiple world champion, and Ironman champion coach.

Tim Guilliams is building Healx at the historic convergence of biology and AI focused on rare diseases.

That is what makes this episode such a strong sequel to Episode 2.

Episode 2 taught us that when reality changes, the best performers do not freeze or thrash. They adapt.

Episode 3 shows that when pain arrives, the best performers do not dramatize it or deny it. They interpret it.

Mirinda’s story captures that perfectly.

In the episode, she describes Kona 2014: coming off the bike around 13.5 to 14 minutes behind, feeling like a failure in that moment, then making a critical internal shift. Instead of obsessing over the outcome, she narrowed the frame: do the best you can in every moment you are in. Stay present. Stay useful. Keep moving. That shift turned pain from panic into process — and eventually into a world title.

That is not just a race story.

It is a model for the brutal middle.

Because in life and startups, people usually make one of two mistakes when pressure rises.

They feel pain and assume they are losing.
Or they feel warning signals and call it toughness.

Both are forms of misreading.

And that is why this conversation feels like the natural continuation of Episode 2’s core idea.

Fluid intelligence is the ability to adapt when the field changes.
Pain literacy is the ability to adapt when the cost becomes real.

One is about reading the environment.
The other is about reading yourself.

The best long-game operators need both.

Ian O’Brien makes that point with the precision of a great coach. His green / amber / red framework is simple, but underneath it is a sophisticated performance principle: pain is not a badge. It is information. Some of it means growth. Some of it means adjust. Some of it means stop. The skill is not becoming numb to pain. The skill is classifying it correctly before ego takes over.

Tim brings the founder version.

His story matters because it avoids the usual startup mythology. He does not argue that burnout is the price of ambition. He shows the opposite. In 2022, after a financing collapse, layoffs, a newborn, and severe strain at home, he kept going but later described himself as deep in the red. What changed was not a slogan. It was a system: sleep, training, recovery, calendar protection, and explicit permission to put his own oxygen mask on first.

That is a much more mature model of performance than the usual “embrace the grind” script.

And the science underneath it is strong.

The deeper science underneath the episode

1) Stress is not only an event. It is an appraisal.
Stanford psychologist Alia Crum and colleagues argue that stress mindset shapes how people respond to stress, and Crum and Jeremy Jamieson later proposed a “stress optimization” approach that integrates mindset and reappraisal. The practical point is simple: the body’s activation may be similar, but the meaning assigned to it changes performance. That maps closely to Mirinda’s Kona shift from “I’m failing” to “What is the best next moment I can execute?”

Evidence snapshot
Threat appraisal → narrower coping and worse execution
Functional appraisal / reappraisal → better coping and stronger performance responses

Authors / source
Alia J. Crum, Stanford University; Jeremy P. Jamieson, then University of Rochester.

2) Endurance is cognitive as well as physical.
Research led by Samuele Marcora found that mental fatigue impairs endurance performance by increasing perceived effort. Later reviews and meta-analyses have reinforced the negative effect of mental fatigue on endurance outcomes. In plain English: how hard something feels, and how well you interpret that feeling, changes what you can do next. That supports both Mirinda’s race account and Ian’s signal-based coaching model.

Evidence snapshot
Higher mental fatigue → higher perceived effort → worse endurance performance

Authors / source
Samuele Marcora and colleagues; Jérémy Habay and colleagues.

3) Elite athletes do not merely tolerate pain more. They often discriminate it better.
A systematic review by Jan Tesarz and colleagues found that athletes generally show higher pain tolerance than normally active controls. More recent work comparing elite and high-level athletes with non-athletes also found increased pain tolerance and different pain processing. The practical takeaway is not that champions ignore pain. It is that trained performers often become more accurate in distinguishing productive discomfort from genuine threat.

Evidence snapshot
Non-athletes → lower tolerance and less practiced pain interpretation
Elite / high-level athletes → higher tolerance and different pain processing

Authors / source
Jan Tesarz et al.; S.D. Pettersen et al.

4) Interoception matters.
Interoception is the sensing of internal bodily signals. A 2022 study found that perceived fatigue limits endurance performance and that interoception plays an important role in regulating sustained physical activity. Related central-fatigue work also links interoception and motivation to endurance regulation. That gives biological support to Mirinda’s feel for the race and Ian’s emphasis on reading the body before damage occurs.

Evidence snapshot
Better internal signal awareness → better effort regulation → better pacing and choices

Authors / source
A. Greenhouse-Tucknott et al.; Tom McMorris et al.

5) Recovery is performance architecture, not indulgence.
This matters especially as a bridge from Episode 3 into the deeper lesson here. Sleep and recovery are not side notes to pain management. They are what stop useful pain from turning into degraded judgment. Stanford’s Rafael Pelayo is a clinical professor in Stanford’s Sleep Medicine division and has emphasized that improving sleep has downstream effects on both mental and physical health. Cheri Mah’s Stanford sleep-extension study in collegiate basketball players found improvements in sprint performance, shooting accuracy, reaction time, mood, and daytime sleepiness, and Stanford has since described her as a pioneer in bringing sleep medicine to athletes to improve performance.

Evidence snapshot
Better sleep / recovery → better mood, reaction time, physical performance
Poor sleep / poor recovery → worse judgment, worse adaptation

Authors / source
Rafael Pelayo, Stanford University; Cheri D. Mah, Stanford-trained sleep researcher and applied elite-performance specialist.

So if Episode 2 gave us a framework for adapting to external change, Episode 3 gives us a framework for adapting to internal strain.

That is why these two episodes belong together.

Episode 2: When the environment changes, do not cling to the old plan.
Episode 3: When the cost rises, do not rush to label the pain. Read it first.

That is a powerful sequence for founders and investors.

Because in startups, the same mistake appears over and over again: people confuse intensity with truth.

They assume pain means:
push harder,
change everything,
or collapse.

But the better operators do something calmer.

They ask:

Is this the cost of growth?
Is this a signal to adapt?
Or is this a sign that recovery is overdue?

That is not softness.

That is judgment.

And that, more than raw grit, is usually what separates growth from burnout.

The difference between growth and burnout is often one skill: reading pain correctly.

So the real lesson from Mirinda Carfrae is not that champions can suffer more.

It is that champions become more literate in suffering.

They learn how to interpret the signal.
They learn when to stay with it.
They learn when to change course.
And they learn when recovery is part of the work.

That is the deeper follow-up to Episode 2.
And it is the deeper meaning underneath Episode 3.

If you build, invest, or operate in long uncertain cycles, that is not just a sports lesson.

It is an operating advantage.

Disclaimer

Endurance Capital is intended for educational and informational purposes only. Nothing in this article, episode, or accompanying materials should be construed as investment, legal, medical, or professional advice. Views expressed by guests are their own. Medical and scientific references are included for context and should not replace individualized professional guidance.

Editorial note

This article was produced as a follow-up to Episode 2 and to the Episode 3 announcement published yesterday. It draws on the original Episode 3 conversation with Mirinda Carfrae, Ian O’Brien, and Tim Guilliams, together with supporting research from reputable academic and institutional sources. It was refined with the assistance of AI as an editorial tool; the framing, conclusions, and final wording reflect my own judgment.

If Episode 2 was about adapting when the middle stops making sense, Episode 3 is about something even harder:

what to do when the middle starts to hurt.

That is what this Endurance Capital conversation with Mirinda Carfrae made clear.

Follow Episode 3 | Pain with Purpose | Pain to Performance here:
🎥 Youtube → Episode 3 | Pain to Performance

🎧 Listen on Spotify, Apple Podcasts, Amazon Music and Deezer.

When people talk about pain in elite performance, they usually go in one of two wrong directions.

They either glorify it.
Or they treat it as failure.

But that is not what stood out most in Rinny.

What stood out was clarity.

Not denial.
Not brute-force toughness.
Not a romantic attachment to suffering.

A more useful relationship with pain:

the ability to read it, work with it, and keep turning it into forward motion — without letting it become damage.

Rinny describes it through racing, not theory.

Coming out of the water behind.
Still chasing on the bike.
And then the marathon — where the real race begins only after the body has already started negotiating.

Again and again, the pattern is the same:

The best performers do not panic when pain arrives.
They do not make it dramatic.
They do not confuse discomfort with defeat.

They stay in it long enough to understand what it means.

That is what makes this conversation relevant far beyond triathlon.

In startups, the equivalent is obvious.

The milestone slips.
The product works, but not yet well enough.
The validation takes longer than expected.
Progress becomes invisible before it becomes real.

Tim Guilliams, founder of Healx, puts it simply:

In deep tech, you can be doing the right work and still feel like nothing is moving.

At that point, the challenge is not effort.

It is holding clarity without immediate proof.

Ian O’Brien makes the same point from the coaching side.

Pain is not a badge.
It is information.

Some pain means you are exactly where growth happens.
Some pain means your system is starting to break.
Some pain means your judgment is about to degrade.

The difference is everything.

Because most ambitious people only have two modes:

push harder
or stop

What they lack is a third:

interpret before reacting.

That is what this episode is really about.

Not suffering for its own sake.
Not resilience as identity.
Not “embracing the grind.”

Something more practical:

how to make pain useful.

Rinny’s version is race composure.
Ian’s version is trained judgment.
Tim’s version is founder endurance.

All three point to the same rule:

When pain arrives, do not rush to label it.
Read it first.

This is where founders tend to make expensive mistakes.

Under pressure, they either:

change direction too early — reacting to discomfort
or
push too long — mistaking damage for discipline

Both come from the same place:

losing the ability to read clearly.

The better operators do something quieter.

They build systems around the pain.

Recovery rhythms.
Decision limits.
Support structures.
Red lines.
Ways to check whether the current load is building capacity — or eroding it.

That is also why this episode matters for investing.

The long-game operator is not the one who can take the most pain.

It is the one who can stay clear inside it.

They know when to continue.
They know when to adjust.
They know when stopping is not weakness, but preservation.

That is a much better signal than intensity.

Three practical takeaways

1. Do not glamorize pain. Classify it.
Not all pain is useful. Some builds capacity. Some signals risk. The skill is not toughness — it is honesty.

2. Build systems for the brutal middle.
When things get hard, you do not rise to intention. You fall back on structure. Recovery, routines, and decision rules matter most when progress feels least visible.

3. Pain needs meaning — or it becomes noise.
People can endure a lot when effort feels directional. What breaks them is not intensity, but lack of signal.

The philosophy behind Endurance Capital

This is the core idea behind Endurance Capital:

the lessons that shape durable champions are often the same ones that shape durable founders and investors.

Not because sport and business are identical.

But because both reward people who can manage:

uncertainty
fatigue
setbacks
and decision-making over long periods of time

Endurance Capital sits at that intersection.

It is not about motivation.

It is about applied thinking under pressure.

Closing

This episode is really about one question:

When pain arrives, what do you do with it?

That is where serious performers separate from reactive ones.

And where durable builders separate from people who confuse intensity with strength.

Episode 3 of Endurance Capital is built around that distinction.

If you build, invest, or operate in long uncertain cycles, start here.

Disclaimer

Endurance Capital is intended for educational and informational purposes only. Nothing in this episode or accompanying materials should be construed as investment, legal, medical, or professional advice. Views expressed by guests are their own.

Editorial note

This article was produced from original interviews and source material and refined with the assistance of AI as an editorial tool. The ideas, conclusions, and final wording reflect my own judgment.

If Episode 1 was about pacing for an impossible race, Episode 2 is about something even harder:

adapting when the middle stops making sense.

That is what this Endurance Capital conversation with Mario Mola made clear.

Follow Episode 2 | Fluid intelligence here:
🎥 Youtube → Endurance Capital | Episode 2 | Fluid Intelligence

🎧 Listen on Spotify, Apple Podcasts and Amazon Music.

When people think about elite performance, they often picture intensity, aggression, and force.

But that is not what stood out most in Mario.

What stood out was clarity.

Not abstract intelligence.
Not overthinking.
Not control for its own sake.

A more useful kind of intelligence:

fluid intelligence — the ability to stay composed, read the field, and make the next best move while everything is moving.

Mario describes it through experience, not theory.

A flat tire in a race that should have ended his chances.
A swim exit 30 to 40 seconds behind the lead group.
A split-second choice between going all-in alone or waiting, regrouping, and preserving the race.

Again and again, the pattern is the same:

The best performers do not waste energy fighting the fact that reality changed.
They absorb the change, process it, and act.

That is what makes this conversation relevant far beyond triathlon.

In startups, the equivalent is obvious.

Markets shift.
Customer behavior surprises you.
Pricing assumptions break.
A launch you thought was imminent suddenly stalls.

Roberto Iannone, co-founder of Zoundream.com put it well: in deep tech, you can spend years building toward what looks like a clear market entry point, only to discover that the business model, pricing, or customer behavior is not what you thought. At that point, the skill is not sticking to the original roadmap out of ego. The skill is learning fast enough to adapt without losing conviction.

Elite coach Ian O’Brien makes the same point from the coaching side.

The goal is not perfection.
The goal is not executing a plan that never changes.
The goal is developing people who can stay in the moment, regain control quickly, and make better decisions under pressure.

His line stayed with me:

No plan survives contact with the enemy.

The edge goes to the person who can adapt without becoming emotional, rigid, or performative.

That applies to founders as much as athletes.

A founder under pressure often makes one of two mistakes:

They freeze.
Or they thrash.

Both come from the same place: losing the ability to read clearly.

Fluid intelligence is the alternative.

It is not passivity.
It is not softness.
It is not “going with the flow.”

It is calm, active adaptation.

Mario’s version is race intelligence.
Ian’s version is trained composure.
Roberto’s version is founder resilience.

All three point to the same rule:

When the environment changes, do not rush to defend your old plan.
Get clear enough to make the next good decision.

That is also why this episode matters for investing.

The long-game operator is not the one who predicts every turn perfectly.
It is the one who sees what is actually happening sooner, adjusts without drama, and preserves the ability to act with precision.

Three practical takeaways

1. Put yourself under pressure deliberately.
Mario’s advice is simple: if you want to think more clearly under pressure, practice making decisions under pressure. That is how composure becomes available when you need it.

2. Train adaptation, not just execution.
Ian’s view is that this is trainable. Decision reviews, simulations, reacting to surges, changing instructions mid-session — all of it builds autonomous decision-makers rather than people who only perform when conditions stay clean.

3. Do not romanticize 24/7 founder grind.
Roberto is clear that constant cognitive overload is not a badge of seriousness. You need enough distance to think. Sometimes the clearest answer arrives after sleep, space, or simply talking the problem out loud.

The philosophy behind Endurance Capital

This is the core idea behind Endurance Capital:

the lessons that create durable champions are often the same ones that create durable founders, investors, and operators.

Not because sport and business are identical. They are not.

But because both reward people who can manage energy, uncertainty, setbacks, timing, and decision-making over long periods of time.

Endurance Capital sits at that intersection.

It is a place to study elite performance not as entertainment, but as applied thinking.

What does pacing teach us about capital allocation?
What does composure teach us about leadership?
What does biological resilience teach us about the long game?

That is the lens.

Not hustle mythology.
Not motivational noise.
Not borrowed prestige from elite sport.

The goal is to extract principles that are actually useful for people building serious things.

Closing

This episode is really about one question:

What do you do when reality stops matching the script?

That is where champions separate from winners.
And where durable founders separate from reactive ones.

Episode 2 of Endurance Capital is now live with Mario Mola, Ian O’Brien, and Roberto Iannone.

If you build, invest, or operate in uncertain environments, start here.

Disclaimer

Endurance Capital is intended for educational and informational purposes only. Nothing in this episode or accompanying materials should be construed as investment, legal, medical, or professional advice. Views expressed by guests are their own.

Editorial note

This article was produced from the original interviews and source material and refined with the assistance of AI as an editorial tool. The ideas, conclusions, and final wording reflect my own judgment.