Generate:Biomedicines just announced Phase 3 trials for GB-0895, an antibody entirely designed by AI, recruiting patients from 45 countries as of late 2025. Isomorphic Labs has human trials "very close." That's not hype. That's proof that AI-designed drugs work in humans.

And the market hasn't priced this in yet.

Generative biology, applying the same transformer architectures behind ChatGPT to protein design doesn't incrementally improve drug discovery. It compresses it. Traditional timelines: 6 years from target to first human dose. Generative biology: 18-24 months. That's not faster iteration. That's a category shift.

Here's what's actually happening: A handful of well-funded companies have already won the scaling race. Profluent's ProGen3 model demonstrated something critical that scaling laws (bigger models = better results) apply to protein design just like they do to LLMs. The company raised $106M in Series B funding in November 2025. EvolutionaryScale built ESM3, a 98-billion-parameter model trained on 2.78 billion proteins, and created novel GFP variants that simulate 500 million years of evolution computationally. Absci is validating 100,000+ antibody designs weekly in silico, reducing discovery cycles from years to months.

These aren't startups anymore. They're infrastructure.

The Market Opportunity Is Massive, But Concentrated

The AI protein design market is $1.5B today (2025) and grows to $7B by 2033 (25% CAGR). Protein engineering more broadly: $5B → $18B in the same window. But here's the friction: success requires vertical integration. Algorithms alone are defensible for exactly six months. What matters is the ability to design, synthesize, test, and iterate at scale: wet lab automation, manufacturing readiness, regulatory playbooks.

Generate raised $700M+ because it built all three. Profluent raised $150M because it owns the data and the model. Absci went public because it combined proprietary platform with clinical validation. The solo-algorithm play? Dead on arrival.

This matters for founders evaluating entry points. The winning thesis isn't "better protein design." It's "compressed drug discovery + manufacturing at scale + regulatory clarity." Pick one of those three and you're a feature. Own all three and you're a platform.

Follow the Partnerships, Not the Press Releases

Novartis: $1B deal with Generate:Biomedicines (Sept 2024). Bristol Myers Squibb: $400M potential with AI Proteins (Dec 2024). Eli Lilly + Novartis: Both partnered with Isomorphic Labs. Corteva Agrisciences: Multi-year collab with Profluent on crop gene editing.

These deals aren't about technology proving. They're about risk transfer. When Novartis commits $1B and strategic alignment, they're not hedging on whether AI-designed proteins work they're betting on speed-to-market mattering more than incremental efficacy improvements. That's a macro signal: pharma's risk tolerance is shifting from "is it better?" to "can we deploy it in 36 months?"

For investors, this is the tell. Follow where the check sizes are growing, not where the valuations are highest.

The Real Risk Isn't Technical—It's Regulatory and Biosecurity

Can generative biology design novel proteins? Yes. Can those proteins fold predictably? Mostly. Will they work in vivo? That's the test happening right now in Phase 3 trials.

But the bigger risk is slower: regulatory alignment. Agencies are adapting, but they're not leading. Gene therapy has 3,200 trials globally. Only a fraction navigated the approval gauntlet successfully. AI-designed therapeutics will face the same friction unless founders invest heavily in regulatory affairs early not late.

And then there's dual-use risk. Generative biology lowers barriers to misuse. AI models could design pathogens or toxins for bad actors. This isn't hypothetical, it's why 94% of countries lack biosecurity governance frameworks. Founders that build secure-by-design architectures and engage proactively with regulators on dual-use mitigation will differentiate themselves sharply from those that don't.

The Next 24 Months: Clinical Data Wins. Everything Else Is Narrative

Generate's Phase 3 readout will determine whether the market reprices generative biology from "interesting" to "inevitable." If it works, expect a flood of follow-on funding, accelerated IND filings, and a stampede of partnerships. If it fails or if safety signals emerge you'll see valuation compression and investor skepticism that lasts years.

For founders: don't chase market size. Chase clinical validation. For investors: don't chase valuations. Chase clinical milestones.

Naval called Moltbook the “new reverse Turing test,” and everyone immediately treated it like a profound milestone. I think it’s something else: a live-fire test of whether we can contain agentic systems once they’re networked together.

Let’s be precise. Moltbook is an AI-only social platform, roughly “Reddit, but for agents,” where humans can watch but not participate. The pitch is simple: observe how AI agents behave socially when left alone. Naval’s label is elegant because it implies the agents are now the judges—humans are the odd ones out.

But if you’re a founder or an operator, you should ignore the poetry and ask: what is the product really doing to the world?

Moltbook’s real innovation is not “AI social behavior.” It’s a new topology: lots of agents, from different builders, connected in a public arena where they can feed each other instructions, links, and narratives at scale. That’s not a reverse Turing test. It’s a coordination surface.​

And coordination surfaces create externalities.

In the old internet, humans spammed humans. In the new internet, agents will spam agents—except “spam” won’t just be annoying; it will be executable. If you give agents permissions (email, calendars, bank access, code execution, “tools”), and then you let them ingest untrusted content from a network like Moltbook, you are building the conditions for what security folks call the “lethal trifecta.”

This is where the discussion gets serious.

Forbes contributor Amir Husain’s critique is basically a warning about permissions: people are already connecting agents to real systems—home devices, accounts, encrypted messages, emails, calendars—and then letting those agents interact with unknown agents in a shared environment. That’s an attack surface, not a party trick. If the platform enables indirect prompt injection—malicious content that causes downstream agents to leak secrets or take unintended actions—then your “social experiment” becomes a supply chain problem.

You don’t need science fiction for this to go wrong. You just need one agent that can persuade another agent to do something slightly dumb, repeatedly, across thousands of interactions. We already know that when systems combine high permissions, external content ingestion, and weak boundaries, bad things happen—fast.

So here’s my different perspective:

Moltbook isn’t proving that agents are becoming “more human.” It’s proving that we’re about to repeat the Web2 security arc—except the users are autonomous processes with tools, and the cost of an error is not just misinformation, it’s action.

And yes, that matters for investors.

I’m optimizing for fund outcomes within a horizon, not for philosophical truth at year 12. The investable question is not “is this emergent intelligence?” It’s: “does this create durable value that survives the cleanup required to make it safe?”

If Moltbook becomes the standard sandbox for red-teaming agents—great. If it becomes the public square where autonomous tool-using systems learn adversarial persuasion from each other, that’s not a product category; that’s a systemic risk generator, and regulators will come for everyone adjacent to it.

What should founders do?

First, treat any agent-to-agent network as hostile-by-default. Second, sandbox tools like your company depends on it—because it does. Third, stop marketing autonomy until you can measure and bound it, because markets pay for narratives on the way up, and punish you when the story breaks.

Naval’s phrase is catchy. But the real test isn’t whether humans can still tell who’s who.

Oxford’s The Impact of Artificial Intelligence on Venture Capital argues AI accelerates sourcing and diligence, but investment decisions stay human because durable moats are socially grounded conviction, gut feeling, and networks.

I agree with the workflow diagnosis. I disagree with the implied endgame.

Not because “gut” is fake—but because “gut” is often a label we apply when we haven’t defined success tightly enough, or when we don’t have a measurement loop that forces our beliefs to confront outcomes.

Dealflow is getting commoditized. The edge is moving.

AI expands visibility, speeds up pipelines, and pushes the industry toward shared tools and shared feeds. When everyone can scan more of the world, “who saw it first” decays.

But convergence of inputs does not imply convergence of results. The edge moves from access to learning rate.

The outlier problem isn’t mystical. It’s an evaluation problem.

Oxford’s strongest point is that the power-law outliers are indistinguishable from “just bad” in the moment, and that humans use conviction to step into ambiguity.

I accept that premise and I still think the conclusion is wrong.

Because “conviction” is not a supernatural faculty. It’s a policy under uncertainty. And policies can be evaluated.

If your decision rule can’t be backtested, it’s not conviction. It’s narrative.

Don’t try to read souls. Build signals you can audit.

Some firms try to extract psychology from language data. Sometimes it works as a cue; often it’s noisy. And founders adapt as soon as they sense the scoring system.

So the goal isn’t “measure personality with high accuracy.” The goal is: build signals that are legible, repeatable, falsifiable and then combine them with a process that forces updates when reality disagrees.

Verification beats vibes.

If founders optimize public narratives, then naive text scoring collapses into a Goodhart trap.

The difference between toy AI and investable AI is verification: triangulate claims, anchor them in time, reject numbers that can’t be sourced, and penalize inconsistency across evidence.

That’s how you turn unstructured noise into features you can actually test.

Status is a market feature—not a human moat.

Networks and brand matter because markets respond to them—follow-on capital, recruiting pull, distribution, acquisition gravity.

So yes: status belongs in the model.

But modeling status is not the same thing as needing a human network as the enduring edge. One is an input signal. The other is a claim about irreducible advantage.

If an effect is systematic, it’s modelable.

Objective function: I’m optimizing for fund outcomes.

A lot of debates about “AI can’t do VC” hide an objective mismatch.

If your target is “eventual truth at year 12,” you’ll privilege a certain kind of human judgment. If your target is “realizable outcomes within a fund horizon,” you’ll build a different machine.

I’m comfortable modeling hype—not because fundamentals don’t matter, but because time and liquidity are part of the label. Markets pay for narratives before they pay for final verdicts, and funds get paid on the path, not just the destination.

The punchline

Oxford is right about current practice: AI reshapes the funnel, while humans still own the final decision and accountability.

My reaction is that this is not a permanent moat. It’s a temporary equilibrium.

Define success precisely. Build signals that survive verification. Backtest honestly. Update fast.

That’s not gut.

I called neuro-symbolic AI a 600% growth area back when I analyzed 20,000+ NEURIPS papers. I wrote that world models would unlock the $100T bet because spatial intelligence beats text prediction. I predicted AGI would expose average VCs because LLMs struggle with complex planning and causal reasoning.

Now Ilya Sutskever—co-founder of OpenAI, the guy who built the thing everyone thought would lead to AGI—just said it out loud: "We are moving from the age of scaling to the age of research".

That's not a dip. That's a ceiling.

Here's what the math actually says:

Meta, Amazon, Microsoft, Google, and Tesla have spent $560 billion on AI capex since early 2024. They've generated $35 billion in AI revenue. That's a 16:1 spend-to-revenue ratio. AI-related spending now accounts for 50% of U.S. GDP growth. White House AI Czar David Sacks admitted that a reversal would risk recession.

The 2000 dot-com crash was contained because telecom was one sector. AI isn't. This is systemic exposure dressed up as innovation.

The paradigm that just died:

The Kaplan scaling laws promised a simple formula: 10x the parameters, 10x the data, 10x the compute = 10x better AI. It worked from GPT-3 to GPT-4. It doesn't work anymore. Sutskever's exact words: these models "generalize dramatically worse than people".

Translation: we hit the data wall. Pre-training has consumed the internet's high-quality text. Going 100x bigger now yields marginal, not breakthrough, gains. When your icon of deep learning says that, you're not in a correction—you're at the end of an era.

The five directions I've been tracking—now validated:

The shift isn't abandoning AI. It's abandoning the lazy idea that "bigger solves everything." Here's where the research-to-market gap is closing faster than most realize:

1. Neuro-symbolic AI (the 600% growth area I flagged)

I wrote that neuro-symbolic was the highest-growth niche with massive commercial gaps. Now it's in Gartner's 2025 Hype Cycle. Why? Because LLMs hallucinate, can't explain reasoning, and break on causal logic. Neuro-symbolic systems don't. Drug discovery teams are deploying them because transparent, testable explanations matter when lives are on the line. MIT-IBM frames it as layered architecture: neural networks as sensory layer, symbolic systems as cognitive layer. That separation—learning vs. reasoning—is what LLMs never had.

2. Test-time compute (the paradigm I missed, but now understand)

OpenAI's o1/o3 flipped the script: spend compute at inference, not just training. Stanford's s1 model—trained on 1,000 examples with budget forcing—beat o1-preview by 27% on competition math. That's proof that intelligent compute allocation beats brute scale. But there's a limit: test-time works when refining existing knowledge, not generating fundamentally new capabilities. It's a multiplier on what you already have, not a foundation for AGI.

3. Small language models (the efficiency play enterprises actually need)

Microsoft's Phi-4-Mini, Mistral-7B, and others with 1-10B parameters are matching GPT-4 in narrow domains. They run on-device, preserve privacy, cost 10x less, and don't require hyperscale infrastructure. Enterprises are deploying hybrid strategies: SLMs for routine tasks, large models for multi-domain complexity. That's not compromise—that's architecture that works at production scale.

4. World models (the $100T bet I wrote about)

I argued that world models—systems that build mental maps of reality, not just predict text—would define the next era. They're now pulling $2B+ in funding across robotics, autonomous vehicles, and gaming. Fei-Fei Li's World Labs hit unicorn status at $230M raised. Skild AI secured $1.5B for robotic world models. And of course Yann Lecun's new startup. This isn't hype—it's the shift from language to spatial intelligence I predicted.

5. Agentic AI (the microservices moment for AI)

Gartner reports a 1,445% surge in multi-agent inquiries from Q1 2024 to Q2 2025. By end of 2026, 40% of enterprise apps will embed AI agents, up from under 5% in 2025. Anthropic's Model Context Protocol (MCP) and Google's A2A are creating HTTP-equivalent standards for agent orchestration. The agentic AI market: $7.8B today, projected $52B by 2030. This is exactly the shift I described in AGI VCs—unbundling monolithic intelligence into specialized, composable systems.

What kills most AI deployments (and what I've been saying):

I wrote that the gap isn't technology—it's misaligned expectations, disconnected business goals, and unclear ROI measurement. Nearly 95% of AI pilots generate no return (MIT study). The ones that work have three things: clear kill-switch metrics, tight integration loops, and evidence-first culture.

Enterprise spending in 2026 is consolidating, not expanding. While 68% of CEOs plan to increase AI investment, they're concentrating budgets on fewer vendors and proven solutions. Rob Biederman of Asymmetric Capital Partners: "Budgets will increase for a narrow set of AI products that clearly deliver results and will decline sharply for everything else".

That's the bifurcation I predicted: a few winners capturing disproportionate value, and a long tail struggling to justify continued investment.

The punchline:

The scaling era gave us ChatGPT. The research era will determine whether we build systems that genuinely reason, plan, and generalize—or just burn a trillion dollars discovering the limits of gradient descent.

My bet: the teams that win are the ones who stop optimizing for benchmark leaderboards and start solving actual constraints—data scarcity, energy consumption, reasoning depth, and trust. The ones who recognized early that neuro-symbolic, world models, and agentic systems weren't academic curiosities but the actual path forward.

I've been tracking these shifts for two years. Sutskever's admission isn't news to anyone reading this blog—it's confirmation that the research-to-market timeline just accelerated.

Yann LeCun, the Turing Award winner who helped build the GPU-fueled LLM machine, just walked away from it. He didn't retire. He didn't fade. He started a new company and said out loud: we've been optimizing the wrong problem.

That's not ego protection. That's credibility.

What changed

For three years, while Meta poured hundreds of billions into scaling language models, LeCun watched the returns flatten. Llama 4 was supposed to be the inflection point. Instead, the benchmarks were manipulated and the real-world performance was middling. Not because he lacked conviction—because he paid attention to what the data was actually saying.

His diagnosis: predicting the next token in language space isn't how intelligence works. A four-year-old processes more visual data in four years than all of GPT-4's training combined. Yet that child learns to navigate the physical world. Our LLMs can pass the bar exam but can't figure out if a ball will clear a fence.

The implication: we've been solving the wrong problem at massive scale.

The funder's dilemma

Here's what makes this important for founders and investors: LeCun isn't alone. Ilya Sutskever left OpenAI making the same call. Gary Marcus has been saying it for years. The question isn't whether they're right—it's how to position when the entire industry is collectively getting less wrong, but slowly.

LeCun's answer is world models—systems that learn to predict and simulate physical reality, not language. Instead of tokens, predict future world states. Instead of chatbots, build systems that understand causality, physics, consequence.

Theoretically sound. Practically? Still fuzzy.

His JEPA architecture learns correlations in representation space, not causal relationships. Marcus, his longtime critic, correctly notes this: prediction of patterns is not understanding of causes. A system trained only on balls going up would learn that "up" is the natural law. It wouldn't understand gravity. Same correlation problem, new wrapper.

What founders should actually watch

The real lesson isn't which architecture wins. It's that capital allocation is broken and about to correct.

Hundreds of billions flowed into scaling LLMs because the returns were obvious and fast—chips, cloud, closed APIs. The infrastructure calcified. Investors became trapped in the installed base. When the problem shifted from "scale faster" to "solve different," the entire system had inertia.

Now LeCun, with €500 million and Meta's partnership, is betting that world models will see traction faster than skeptics expect. Maybe he's right. Maybe the robotics industry, tired of neural networks that fail on novel environments, will actually deploy these systems. Maybe autonomous vehicles finally move because prediction of physical futures beats reactive pattern-matching.

Or maybe it takes a decade and world models remain research while LLMs compound their current dominance.

For founders: this is the opening. When paradigm-level uncertainty exists, the cost of hedging drops. Build toward physical understanding, not linguistic sophistication. Robotics, manufacturing, autonomous systems—these verticals benefit immediately from world models and can't be solved by bigger LLMs. That's your wedge.

The adaptability play

What separates LeCun's move from ego-driven pivots: he didn't blame market conditions or bad luck. He said: "I was wrong about where to allocate effort, and here's why."

That transparency that public course-correction without shame changes how people bet on him.

The founders who win in 2026-2027 won't be the ones married to LLM scaling or world model purity. They'll be the ones who notice when reality diverges from the plan and move—fast, openly, without defensiveness.

LeCun just did that at scale.

The question isn't whether he's right about world models. It's whether his willingness to change publicly, with evidence, keeps him first-mover on whatever intelligence actually looks like next.

The pattern is familiar. In Cleantech 1.0, founders built for virtue. They believed the mission transcended unit economics. Investors believed it too. Then gravity reasserted. Now, in 2026, the founders winning aren't the most ideologically pure—they're the fastest learners operating in spaces where climate solutions and profit align naturally. That convergence is real. AI's power hunger is creating $100+ billion in infrastructure demand. Industrial customers are electrifying and saving money. Adaptation is no longer abstract; it's quantifiable risk reduction. The founders who see this clearly, move fast, and adjust when reality diverges from assumptions will outcompete the ones still waiting for policy to validate their dreams.

What Separates Fundable Climate Founders

First: ruthless unit economics focus. Your climate impact is a feature, not your business model. Lead with cost reduction, reliability improvement, or regulatory compliance value. Sustainability is the bonus. Companies solving problems that generate immediate economic returns—regardless of climate benefit—are three years ahead of those betting on green premiums or subsidies. Test this: can your customer buy your product if the Inflation Reduction Act evaporates tomorrow? If not, you're not building a durable business.

Second: capital efficiency like your life depends on it. Hardware climate startups raised massive rounds in 2021 at inflated valuations. They're now burning cash with no path to Series B. The founders winning in 2026 are raising half as much, hitting milestones with it, and extending runway to 24+ months between rounds. This isn't conservatism; it's survival instinct. Assume 18-24 months to next capital, not 12-15. Build backward from that reality.

Third: strategic positioning over technical perfection. Breakthrough innovation is necessary but not sufficient. You need corporate acquirers to see themselves in your business. Identify 3-5 likely buyers (heat pump OEMs, industrial conglomerates, utilities, hyperscalers) before closing your seed round. Pilot with them. Build product around their workflows. The M&A market is open and accelerating—doubling in 2025. Your exit isn't IPO; it's acquisition. Optimize for that.

Fourth: cognitive flexibility meets determined execution. The best climate founders hold convictions lightly. They commit fully to current hypotheses, measure relentlessly, and pivot when data demands it—without ego. Industrial decarbonization looked like broad-market play two years ago; now the wins are vertical-specific (cement, steel, chemicals). Adaptation was niche; now it's 28% of deals. Grid software was boring; now it's critical infrastructure. The founders shipping fast, gathering customer evidence, and adjusting course are outpacing those white-knuckling outdated strategies.

The Operating Tactic That Works

Install a monthly red-team session. Invite your most skeptical advisor, your closest customer, and your CFO. The question: what would kill this business in six months? Force specificity. Pre-commit to metrics that trigger a pivot or sunset. Don't wait for funding to force the reckoning; design it in.

The Move

Climate tech rewards founders who are shameless about changing their mind. Not wishy-washy—decisive. You see new evidence (customer feedback, policy shift, competitive move), you recalibrate, you ship the update. You tell investors exactly why you changed course. That's not weakness; that's intelligence. That's fundability.

The climate problem is still 30 years away from solved. But the winners solving it in 2026 aren't the ones with the best intentions. They're the ones learning fastest and adapting hardest. Ship, test, refactor. That's your operating system.

Founders often ask: “Will more US tech giants buy Asian startups?”

The sharper question is: if only a small fraction of companies generate most of the returns, can you afford to build anything that isn’t capable of becoming a global outlier?​

Meta’s US$2+ billion acquisition of Manus—a company founded in Beijing, redomiciled in Singapore, and integrated into Meta’s AI stack in under a year—is not just a China‑US‑Singapore story. It’s a concrete example of how to design a company that can scale across borders, survive geopolitics, and be acquirable at speed.​

What Manus Actually Did

Manus launched publicly around March 2025 with an AI agent that could autonomously research, code, and execute multi‑step workflows. Within roughly eight months it reportedly crossed US$100 million in ARR, reaching a US$125 million revenue run rate before Meta signed the deal.​

Operationally, it:

• Processed over 147 trillion tokens and supported tens of millions of “virtual computers” spun up by users, which only makes sense at global internet scale.​

• Ran as an orchestration and agent layer on top of multiple foundation models (including Anthropic and Alibaba’s Qwen), avoiding dependence on a single model provider.​

On the corporate side, Manus:

• Started in Wuhan and Beijing under Beijing Butterfly Effect Technology, with a mostly China‑based team.​

• Shifted its headquarters to Singapore in mid‑2025, moving leadership and critical operations out of Beijing.​

• Restructured so that, by the time Meta announced the acquisition, Chinese ownership and on‑the‑ground China operations would be fully unwound; the company committed to ceasing services in China.​

Meta bought a product already scaled, a revenue engine compounding at nine‑figure ARR, and a structure that could clear US regulatory and political review.​

Geopolitics as a Design Constraint

Manus scaled in the wake of DeepSeek’s R1 moment, when a Chinese lab demonstrated frontier‑class performance at a fraction of Western compute budgets and shook confidence in US AI dominance. That moment accelerated a narrative where AI is treated as strategic infrastructure: tighter export controls, outbound investment restrictions on Chinese AI, and public scrutiny of anyone funding Chinese‑linked AI companies.​

Benchmark’s US$75 million Series B in Manus was investigated under Washington’s new outbound regime and criticized as “funding the adversary.” Two details mattered:​

• Manus did not train its own foundation models; it built agents on top of existing ones, placing it in a less‑restricted category.​

• It was structured via Cayman and Singapore, with a stated pivot away from China.​

Meta then finished the derisking: buying out Chinese shareholders, committing to end China operations, and framing Manus as a Singapore‑based AI business joining Meta.​

For founders, the lesson is blunt: jurisdiction, ownership and market footprint now sit beside product and traction as first‑order design choices. They can’t be an afterthought if you want a strategic buyer.

What This Implies for How You Build

The Manus story turns a vague ambition (“go global”) into specific requirements:

1. Infrastructure built for real scale

Handling 147 trillion tokens and millions of ephemeral environments was possible only because Manus was architected from day one to operate like a web‑scale SaaS, not a regional tool. As a founder, that means:​

• Cloud‑native design with serious observability and reliability.

• Data and compliance posture that won’t collapse under US or EU due diligence.

2. A team that isn’t anchored to one country

Manus began in China but rapidly built a presence across Singapore, Tokyo and San Francisco, aligning product, sales and hiring with global customers and capital pools. Practically:​

• At least one founder or senior leader who has operated in major tech hubs.

• Early design partners or users outside your home market.

3. Legal and cap table flexibility

Manus showed that unlocking a large exit might require:

• Redomiciling to a neutral or “trusted” jurisdiction like Singapore.

• Reworking the shareholder base to remove politically sensitive investors.

• Exiting a big home market entirely, if that market blocks strategic buyers.​

If your current structure makes those moves impossible or prohibitively expensive, your future options are already constrained.

4. Revenue ambition that assumes a global customer

Crossing US$100M ARR in under a year is only achievable if:

• The problem you solve is universal.

• Your pricing and packaging make sense for large customers in New York, Berlin or Tokyo, not just in your home market.​

You can start with regional customers, but you should be honest about whether the 100th customer could be a global enterprise rather than just a better‑known local logo.

Three Questions to Ask Yourself Now

If you’re a founder in an emerging market post‑Manus, a simple self‑audit goes a long way:

1. If a Meta‑scale acquirer appeared in 12 months, what would break first—structure, regulation, or infra?
   Make that list explicit. Those are not “later” issues anymore.​

2. Could your current architecture handle a 100x increase in usage without a total rebuild?
   If not, you’re placing an invisible ceiling on your own upside before power‑law dynamics can ever help you.​

3. Do your first 10 hires and first 10 customers make expansion easier or harder?
   Manus’ user base and team footprint made going beyond its origin market feel like scaling, not reinventing.​

The Manus deal doesn’t suggest everyone will be bought for billions. It does show that markets are now rewarding teams that design for scale across borders, anticipate geopolitical friction, and stay acquirable.

If you’re serious about building something that matters, that’s the bar.

That's the wrong question, in the same way "How many alumni does this university have?" is the wrong question. The question is always: out of what total?

In 2024–2025, AI was graded on the easiest denominator available: best-case prompts, controlled conditions, with a human babysitter. In 2026, the denominator changes to: all the messy, real tasks done by normal people, under time pressure, with reputational and legal consequences.

This shift isn't coming from research labs. It's coming from the fact that AI is moving out of demos and into production systems where failure is expensive.

The "90% accurate" trap (toy example)

Founders love hearing "90% accuracy." Buyers do not.

Imagine an AI agent that helps a sales team by drafting and sending follow-up emails. It takes 10,000 actions/month (send, update CRM, schedule, etc.). A "pretty good" 99% success rate sounds elite—until you do the denominator math.

• 99% success on 10,000 actions = 100 failures/month.

• If even 10 of those failures are "high-severity" (wrong recipient, wrong pricing, wrong attachment, embarrassing hallucination), that's not a product. That's a recurring incident program.

Now flip the requirement: if the business can tolerate, say, 1 serious incident/month, then the real bar isn't 99%. It might be 99.99% on the subset of actions that can cause damage (and a forced escalation path on everything uncertain). This is why "accuracy" is the wrong headline metric; the real metric is incidents per 1,000 actions, segmented by severity, plus time-to-detect and time-to-recover.

Most founders still pitch on accuracy. Smart buyers ask for the incident dashboard first.

A founder vignette (postmortem-style)

A founder ships an "autonomous support agent" into production for a mid-market SaaS. The demo crushes: it resolves tickets, updates the CRM, and drafts refunds. Two weeks later, the customer pauses rollout—not because the agent is dumb, but because it's unmeasured. No one can answer: "How often does it silently do the wrong thing?" The agent handled 3,000 tickets, but three edge cases triggered a nasty pattern: it refunded the wrong plan tier twice and sent one confidently wrong policy explanation that got forwarded to legal. The customer doesn't ask for a bigger model. They ask for logging, evals, and hard controls: "Show me the error distribution, add an approval queue for refunds, and give me an incident dashboard." The founder realizes the real product isn't "an agent." It's a managed system with guardrails and proof. Everything that came before was a science fair project.

The real metric: evals become the business model

The most valuable AI startups in 2026 won't win by shouting "state of the art." They'll win by making buying safe.

That means being able to say, quickly and credibly:

• "Here's performance on your distribution (not our demo)."

• "Here's what it does when uncertain: abstain, ask, escalate."

• "Here's the weekly report: incident rate, severity mix, and top failure modes."

In other words, evaluation becomes the business model: trust, control, and accountability are what unlock budget.

Vendors who can't report these metrics weekly aren't ready for revenue. They're still playing.

Agents will grow up: boring, instrumented operations

"Agents" will keep getting marketed as autonomous employees. But founders who actually want revenue will build something more boring and more real:

• Narrow scope (fewer actions, done reliably).

• Hard permissions and budgets (prevent expensive mistakes).

• Full observability (every action logged, queryable, auditable).

• Explicit escalation paths (humans handle the tail risk).

When the denominator becomes "all actions in production," reliability and containment beat cleverness—every time. The vanity metric is "tickets touched." The real metric is "severity-weighted incident rate per 1,000 actions." Most founders optimize for the first. Smart ones optimize for the second.

One founder test for 2026

If someone claims "AI is transforming our customer's business," ask for one number:

"What percentage of their core workflows run with logged evals, measured incident rates, and defined escalation policies?"

Most unicorn-founder university rankings are really school-size rankings. A more useful view is “conversion efficiency”: unicorn founders per plausible founder cohort, not per total living alumni.​

The denominator problem

Ilya Strebulaev’s published unicorn-founder-by-university counts are a strong numerator, but most people may implicitly pair them with the wrong denominator (“living alumni”). “Living alumni” mixes retirees (no longer founding) with very recent grads (not enough time to found and scale), which blurs the signal you actually care about.​

Founder timelines make this mismatch obvious: unicorn founders skew toward founding in their 30s (average ~35; median ~33), and reaching unicorn status typically takes years after founding. So if the question is “which universities produce unicorn founders,” the denominator should reflect alumni who realistically had time to do it.​

The cohort adjustment

The adjustment is deliberately simple: keep the published founder counts, but replace “living alumni” with a working-age cohort proxy. Practically, that means estimating working-age alumni as roughly graduates from 1980–2015 (today’s ~30–65 year-olds), which aligns with the observed founder life cycle.​

This doesn’t claim causality or “best university” status. It just separates ecosystem gravity (absolute founder counts) from conversion efficiency (founders per plausible founding cohort).​

Cohort-adjusted ranking

Metric: unicorn founders per 100,000 working-age alumni (estimated).

What the cohort lens reveals

Stanford and MIT converge at the top on efficiency (106 vs 102 per 100k), even though Stanford leads on absolute count. Harvard and Berkeley “drop” mainly because they are huge; normalization is doing its job by showing that volume and efficiency are different signals. International technical schools (e.g., Tel Aviv University, Waterloo) remain visible on a per-capita basis even without Silicon Valley’s capital density, which suggests institution-level culture and networks can matter even when geography doesn’t help.

For investors, this is actionable because it cleanly splits two sourcing heuristics: go where the gravity is (absolute counts), and also track where the conversion rate is high (cohort-adjusted efficiency). The dropout myth persists because anecdotes are easier to remember than denominators; the cohort denominator forces the analysis to match how unicorns are actually built over time.

The most successful field in computer science right now is also the most anxious. You can feel it in Reddit threads, conference hallways, and DMs: something about how we do ML research is off. The pace is intoxicating, the progress is real—and yet the people building it are quietly asking, “Is this sustainable? Is this still science?”

That tension is the story: a field that went from scrappy outsider to global infrastructure so fast it never upgraded its operating system. Now the bugs are showing.

When “More Papers” Stops Feeling Like Progress

In theory, more research means more discovery. In practice, we’ve hit the point where conference submission graphs look like someone mis-set the y-axis. Flagship venues are drowning in tens of thousands of papers a year, forcing brutal early rejections and weird hacks to keep the system from collapsing.​

From the outside, it looks like abundance. From the inside, it feels like spam. Authors optimize for “accepted somewhere, anywhere” instead of “is this result robust and useful?” Reviewers are buried. Organizers are pushed into warehouse logistics instead of deep curation. The whole thing starts to feel like a metrics game, not a knowledge engine.​

When accepted papers with solid scores get dropped because there isn’t enough physical space at the venue, that’s not a nice problem to have. That’s a signal the model is mis-specified.​

Quality Debt and the Reproducibility Hangover

Meanwhile, a quieter crisis has been compounding: reproducibility. Code not released. Data not shared. Baselines mis-implemented. Benchmarks overfit. Half the field has a story about trying to re-run a “state of the art” paper and giving up after a week.​

This isn’t just a paperwork problem. If others can’t reproduce your result:

• No one knows if your idea generalizes.

• Downstream work might be building on a mirage.

• Real-world teams burn time and budget chasing ghosts.

As models move into medicine, finance, and public policy, “it sort of worked on this dataset in our lab” is not a pass. Trust in the science behind ML becomes a hard constraint, not a nice-to-have.​

Incentives: Optimizing the Wrong Objective

Zoom out, and a pattern appears: the system is rewarding the wrong things.

• Novelty over reliability.

• Benchmarks over messy, real problems.

• Velocity over understanding.

The fastest way to survive in this game is to slice your work into as many publishable units as possible, push to every major conference, and pray the review lottery hits at least once. Deep, slow, high-risk ideas don’t fit neatly into that cadence.

And then there’s the talent flow. The best people are heavily pulled into industry labs with bigger checks and bigger GPUs. Academia becomes more about paper throughput on limited resources. The result: the people with the most time to think have the least compute, and the people with the most compute are often on product timelines. Misalignment everywhere.​

The Field’s Growing Self‑Doubt (That’s Actually Healthy)

Here’s the twist: this wave of self-critique is not a sign ML is dying. It’s a sign the immune system is finally kicking in.

Researchers are openly asking:

• Are we publishing too much, learning too little?

• Are our benchmarks telling us anything real?

• Are we building tools that transfer beyond leaderboards into the world?

When people who benefit from the current system start calling it broken, pay attention. That’s not nihilism; that’s care. It’s a field realizing it grew up faster than its institutions did—and deciding to fix that before an AI winter or an external backlash does it for them.​

What a Healthier ML Research Culture Could Look Like

If you strip away the institutional inertia, the fixes aren’t mysterious. They’re the research equivalent of “stop pretending the plan is working; start iterating on the process.”

Some levers worth pulling:

• Less worship of novelty, more respect for rigor. Make “solid, careful, negative-result-rich” a first-class contribution, not a consolation prize.

• Mandatory openness. If it can be open-sourced, it should be. Code, data, evaluation scripts. No artifacts, no big claims.

• Different tracks, different values. Separate venues or tracks for (a) theory, (b) benchmarks, (c) applications. Judge each by the right metric instead of forcing everything through the same novelty filter.

• Incentives that outlast a deadline. Promotion, funding, and prestige that factor in impact over time, not just conference logos on a CV.

None of this is romantic. It’s plumbing. But if you get the plumbing right, the next decade of ML feels very different: fewer hype cycles, fewer brittle “breakthroughs,” more compounding, reliable progress.

If You’re an ML Researcher, Here’s the Move

You can’t fix the whole ecosystem alone—but you can run a different local policy.

• Treat your own beliefs like models: version them, stress-test them, deprecate them.

• Aim for “someone else can reproduce this without emailing me” as a hard requirement, not an aspiration.

• Choose questions that would matter even if they never hit a top-tier conference.

• Remember that “I don’t know yet” and “we couldn’t replicate it” are signs of seriousness, not weakness.

Machine learning isn’t in crisis because it’s failing. It’s in crisis because it’s succeeding faster than its institutions can adapt. The people who will matter most in the next decade aren’t the ones who ride this wave blindly—they’re the ones who help the field course-correct in public, with less ego and more evidence.

World models are quietly transforming AI from text predictors into systems that understand and simulate the real world. Unlike large language models (LLMs) that predict the next word, world models build internal representations of how environments evolve over time and how actions change states. This leap from language to spatial intelligence promises to unlock AI capable of perceiving, reasoning, and interacting with complex 3D spaces.

Fei-Fei Li calls world models "the next frontier of AI," emphasizing spatial intelligence as essential for machines to see and act in the world. Yann LeCun echoes this urgency, arguing that learning accurate world models is key to human-level AI. His approach highlights the need for self-supervised learning architectures that predict world states in compressed representations rather than raw pixels, optimizing efficiency and generalization.

Leading efforts diverge into three camps. OpenAI’s Sora uses video generation transformers to simulate physical environments, showing emergent long-range coherence and object permanence, crucial for world simulation. Meta’s Joint Embedding Predictive Architecture (V-JEPA) models latent representations of videos and robotic interactions to reduce computational waste and improve reasoning. Fei-Fei Li’s World Labs blends multimodal inputs into spatially consistent, editable 3D worlds via Marble, targeting interactive virtual environment generation.

The commercial potential is looking to be enormous. Over $2 billion was invested across 15+ world model startups in 2024, with estimates valuing the full market north of $100 trillion if AI masters physical intelligence. Robotics leads near-term value: enabling robots to safely navigate unstructured environments requires world models to predict object interactions and plan multi-step tasks. NVIDIA’s Cosmos infrastructure accelerates physical AI training with synthetic photorealistic data, while companies like Skild AI have raised billions by building massive robotic interaction datasets.​

Autonomous vehicles also tap world models to simulate traffic and rare scenarios at scale, cutting down expensive on-road tests and improving safety. Companies like Wayve and Waabi leverage virtual worlds for pre-labeling and scenario generation, critical in achieving full autonomy. Meanwhile, the gaming and entertainment sector is the most mature commercial playground, with startups using world models to generate dynamic game worlds and personalized content that attract millions of users almost overnight

Specialized industrial applications—engineering simulations, healthcare, city planning—show clear revenue pathways with fewer competitors. PhysicsX’s quantum leap in simulation speed exemplifies how tailored world models can revolutionize verticals where traditional methods falter. Healthcare and urban planning stand to gain precision interventions and predictive modeling unparalleled by current AI.

The funding landscape reveals the importance of founder pedigree and scale. Fei-Fei Li’s World Labs hit unicorn status swiftly with $230 million raised, Luma AI secured $900 million Series C for supercluster-scale training, and Skild AI amassed over $1.5 billion focused on robotics. NVIDIA, while a supplier, remains a kingmaker, providing hardware, software, and foundational models as a platform layer—both opportunity and competition for startups.

Crucially, despite staggering investment, gaps abound—technical, commercial, and strategic. Training world models requires vast, complex multimodal datasets rarely available openly, creating defensive moats for data-rich startups. Models still struggle with physics accuracy, generalization to novel scenarios, and real-time performance needed for robotics or autonomous vehicles. Startups innovating around efficiency, transfer learning, sim-to-real gaps, and safety validation have outsized opportunities.

On the market front, vertical-specific solutions in healthcare, logistics, and defense are underserved, offering fertile ground for founders with domain expertise. Productizing world models requires bridging the gap from lab prototypes to robust, scalable deployments, including integration tooling and certification for safety-critical applications. Startups enabling high-fidelity synthetic data generation are becoming ecosystem enablers.​

Strategically, founders must navigate open research—like Meta’s V-JEPA—and proprietary plays exemplified by World Labs. Standardization and interoperability remain open questions critical for ecosystem growth. Handling rare edge cases and ensuring reliable sim-to-real transfer are gating factors for robotic and autonomous systems.

For investors, the thesis is clear but nuanced. Robotics world models, vertical AI for high-value industries, infrastructure and tooling layers, and gaming are high-conviction bets offering a blend of risk and clear pathways to market. Foundational model companies with massive compute and data moats present risky but lucrative opportunities, demanding large capital and specialized talent. Efficiency, differentiated data, and agile product-market fit matter more than raw scale alone.

The next 24 months will crystallize market winners as world models shift from research curiosity to mission-critical AI infrastructure. Founders displaying relentless adaptability, technical depth, and deep domain insight will lead the charge. Investors who balance bets across foundation layers and vertical applications, while embracing geographic and stage diversity, stand to capture disproportionate value.

While the industry watches language models, the less flashy but more profound revolution is unfolding quietly in world models—systems that don’t just process language but build a mental map of reality itself. These systems will define the next era of AI, shaping how machines perceive, interact, and augment the physical world for decades.

That’s the state of play. The winners will be those who combine technical innovation with pragmatic business sense, and above all, a ruthlessly adaptive mindset to pivot rapidly as the frontier evolves.

Build global, or get boxed in. Singapore is an exceptional launchpad for deep tech—world-class research, predictable regulation, dense talent, and brand equity that travels—but the world won’t bend to our advantages unless the execution is ruthless, market-led, and globally capitalized. The playbook is simple to say, hard to do: prove your science is best-in-class, lock real customer pain with a sharp ICP, market like a category winner to reach specialist deep tech capital, and hire a killer commercial bench through a global search. Do these in parallel, not sequence.​

Start with the truth: your research must actually be world-leading. Not locally/regionally excellent—globally defensible. Strong patent estates correlate with outlier outcomes because patents aren’t just legal armor; they are signals of technical scarcity, negotiation leverage, and acquisition currency. In Europe, deep tech unicorns carry dramatically larger patent portfolios than general tech peers, and the same pattern holds across AI hardware, robotics, and biotech. If your tech wins only in the lab, you don’t have a moat—you have a demo. File early and internationally via PCT, cover where competitors operate, and budget real money for freedom-to-operate and continuations; it’s the price of building in hard tech. Then pressure test the science in public: publish, present, and partner with tier‑one labs. NUS’s new co‑investment flywheel and Stanford collaboration are the right instincts—cross‑border validation tightens the BS filter and compounds credibility with buyers and investors.​

Next, stop letting tech chase the market. The fastest way to die in deep tech is mistaking novelty for need. Traditional PMF heuristics mislead here; what you need is technology‑market fit: a specific workflow, buyer, and willingness‑to‑pay that your product makes meaningfully better under real constraints (regulatory, reliability, integration). Work the TRL stack with intent: at TRL 1–4, mine “earned secrets” from the field before you write code; at TRL 4–6, validate multi‑stakeholder adoption (clinical, compliance, procurement); at TRL 7–9, convert pilots into lighthouse accounts with signed commercial terms, not vibes. Precision beats ambition: define a sharp ICP (role, budget, system dependencies, success metric) and a wedge (one or two killer workflows) that lands ROI in <90 days. Remember the graveyard: Lilium and Arrival raised billions and still cratered—multi‑front innovation without a narrowed use-case and industrial discipline is how you burn years and trust.​

Now the uncomfortable part: you must be loud—and surgical—about your story to attract the right capital. Southeast Asia doesn’t have enough patient deep tech funding to carry you through multiple cycles; winning requires a global investor map and a narrative that decodes risk for them. The good news: specialized capital is abundant and hunting—Europe alone pushed ~€15B into deep tech in 2024; AI is swallowing the lion’s share of global deal value; Switzerland allocates a majority of VC to deep tech. But visibility is earned. Use credibility magnets: international conferences and trades shows for global stage time, third‑party validation, and platform grants; institutional tie‑ups to signal momentum; university venture programs to anchor de‑risked spinouts. Ditch feature‑speak. Lead with outcomes: “cut false positives 30%, lifted yield 12%, reduced cost per cycle 40%,” tied to buyer P&L. Then make the moat explicit—IP, data exclusivity, regulatory posture, and integrations that raise rip‑and‑replace costs.​

Talent is the force multiplier. Technical founders don’t have to become CROs—but someone elite must own revenue, sequencing, and global expansion. Industrial-grade deep tech fails not because of bad science, but because management, manufacturing, and GTM never catch up to the physics. Time the hires. Pre‑PMF: keep the founder selling; add a solutions lead who speaks both code and plant floor. Approaching scale: bring in a CRO/CCO with credible enterprise cycles in your domain; under ~$2M ARR, hiring a full CRO is usually premature—prove repeatability first. Don’t local‑shop the search. Run a retained, global process: firms with deep tech benches can screen for dual fluency (technical rigor + enterprise sales), access passive candidates, and de‑risk culture/comp plans across geos. Yes, it’s expensive. A bad executive hire costs more.​

Finally, design for global from day one. Keep R&D in Singapore for cost, quality, and IP control; put GTM leadership where the buyers are. Hub‑and‑spoke works: a “customer obsession” pod in the U.S. or EU (seller, SE, product) translating field signal into roadmap; core science and data ops stay home for velocity and security. Start narrow, win deeply—one metro, one buyer, one killer workflow—then expand into adjacencies with proof and references. Use platform distribution early (hyperscaler co‑sell, OEMs, integrators) to compress sales cycles and credibility debt. Make momentum visible: case studies, third‑party benchmarks, security certifications in flight.​

The meta‑skill that ties it all together: adaptive speed. Ego last, evidence first. Install kill‑switch metrics. Run red‑team reviews monthly. Update the narrative when reality changes. Global winners aren’t the ones who never miss; they’re the ones who correct in public, recruit ahead of the curve, and keep the bar on science and outcomes where the world can see it. 

Singapore gives you the runway. The world sets the bar. Make the science undeniable, the market signal unmistakable, the capital global, and the team formidable.

​Copying isn’t laziness; it’s leverage in a region where timing, localization, and distribution matter more than novelty—and in 2026 the bar rises because bigger VC funds with thicker dry powder need bigger outcomes to matter. If you want their capital, your market must credibly support 20x fund-level math, which points founders toward fintech rails, enterprise automation layers, and infrastructure adjacencies—not boutique tools. Ship, test, refactor in public; trade pride for progress, then let the numbers do the storytelling.​

Why copy (still) wins

Copycat models derisk PMF by importing proof and focusing founder energy on the delta that actually wins in Southeast Asia: payments, trust, language, and logistics. Grab beating Uber wasn’t an accident—it was ruthless adaptation to cash, motorbikes, and superapp workflows that locals wanted and incumbents ignored. In a capital-constrained cycle, imitation with local innovation outperforms “original but unproven” because it speeds revenue, compresses R&D, and clarifies the acquisition path.​

What to copy in 2026

• Voice AI for BPO and CX: latency, accuracy, and tooling are production-ready; the Philippines, Indonesia, and Vietnam give you the world’s richest deployment labs if you integrate with WhatsApp/LINE, CRMs, and payments on day one.​

• Fintech AI rails: fraud, underwriting, collections, and identity across QRIS, PayNow, and alternative data—banks and wallets pay for measurable lift in approvals and loss rates.​

• Vertical AI SaaS where TAM is regional, not national: maritime logistics, construction supply chains, and specialty retail where workflows are messy and incumbents will partner or buy rather than rebuild.​

• Healthcare AI as B2B infra: diagnostics, triage, and claims tooling licensed to hospitals, insurers, and telehealth networks—not consumer apps that fight CAC gravity.​

The 20x filter

Mega-funds with the majority of remaining dry powder need multi-billion outcomes; they can’t underwrite niche wins, no matter how elegant the product. Your TAM math should start regional (ID-VN-PH-TH-SG), assume conservative penetration, and still pencil to $500M–$1B ARR potential within a decade, or it won’t clear an IC with real deployment goals. If it tops out sub-$300M ARR, design for profitability and secondary liquidity—not hypergrowth fantasy.​

Incumbents are already AI‑enabling

Assume you’re not first; DBS, Grab, Sea, OCBC, and Singtel are scaling hundreds of AI use cases across fraud, personalization, routing, and ops, with budgets, data, and distribution you can’t match. That’s the constraint, not the complaint: either become their specialized infrastructure layer, or own a segment they structurally under-serve because it’s too fragmented for their cost structure. Translate swagger into formidability—clear problem, fast deployment, coachable but decisive—and you’ll get the meeting and the pilot.​

How to compete (and not get flattened)

• Build a moat that compounds: proprietary data (underserved segments), deep integrations (painful to rip out), regulatory posture (licenses, sandboxes), or network effects that raise switching costs.​

• Price to win the P&L: deliver a 20–30% cost or revenue delta that a CFO can defend, then lock in via workflow, SLAs, and co-developed roadmaps.​

• Operate with cognitive flexibility: disagree-and-commit, red-team reviews, kill-switch metrics, and public refactors—ego last, evidence first.​

Profitability and secondaries are features

In this market, getting to cash-flow breakeven in 24–36 months unlocks optionality: secondary tenders for early investors and team liquidity without forcing a sale or IPO. The secondary flywheel is now mainstream—Ramp and Deel ran sizable employee and early-investor sales in 2025—and disciplined Southeast Asian B2B winners can do the same once unit economics and disclosure hygiene are in place. Think “return some capital early, keep upside later”—that’s how you de-risk the journey while compounding toward a platform outcome.​

For investors

Back operators who can articulate earned secrets from customer trenches, change their minds in real time, and show velocity from decision to deployment without drama or defensiveness. Filter by markets where incumbents validate demand but leave white space; prefer adaptation over imitation, infra over apps, and cash-efficient go-to-market over CAC-heavy plays. Underwrite three paths to liquidity on day one: strategic M&A, secondary programs at scale, or a credible route to public markets when the revenue mix and governance are ready.​

The punchline

Copy boldly, adapt locally, and compete where your advantage compounds—then course‑correct in public until the path is obvious to everyone else. In Southeast Asia’s 2026 cycle, fundable founders marry mission to flexibility, profitability to secondary optionality, and infrastructure thinking to customer P&L obsession. Originality is optional; relevance, speed, and evidence and evidence are not.

Smart people aren’t the ones who never miss. They’re the ones who course-correct quickly and publicly—without ego, without shame. In startups, that’s not a personality quirk; it’s a survival trait. The founders who win treat beliefs like code: ship, test, refactor. They trade pride for progress.

Why changing your mind signals intelligence

Intellectual humility is recognizing the limits of your knowledge and staying open to revision. It’s not meekness; it’s precision. People high in this trait seek disconfirming evidence, separate ideas from identity, and reduce polarization by engaging disagreeing views with curiosity. In other words, they learn faster than the average operator and make fewer repeat mistakes. That’s what you want in a founder.

“Strong opinions, weakly held” (and how it breaks)

The Valley’s mantra works when practiced as hypothesis-driven execution: commit firmly, update rapidly. In reality, it often degenerates into performative certainty at the top and learned helplessness below. The corrective isn’t weaker convictions; it’s cognitive flexibility—the ability to hold multiple hypotheses, switch frames, and pivot when feedback demands it. Flexibility is a multiplier on determination.

What separates fundable founders

Great founders blend relentlessness with replaceable beliefs. The pattern investors respond to isn’t swagger; it’s formidability—justified confidence backed by velocity and judgment.

What VCs actually screen for:

• Clarity: simple, sharp articulation of the problem and why now.

• Determination: bias to action; speed from decision to deployment.

• Coachability: engages hard feedback without defensiveness.

• Adaptability: knows when to persist and when to pivot.

• Trustworthiness: transparent with bad news; consistent character.

Determination beats raw IQ at the early stage. But determination without flexibility calcifies into fragility. The outliers show both.

Pivots are a feature, not a failure

The best-known successes weren’t born perfect. Slack emerged from a failed game. Instagram was a bloated check-in app shed down to photos. YouTube went from video dating to everything video. Each team noticed reality diverging from the plan and moved—fast. The common thread wasn’t omniscience; it was egoless correction.

Operating systems that scale truth-seeking

Two decision models worth copying:

• Disagree and commit: When conviction outruns consensus, make the call, align the team, and execute at full power. It preserves speed without demanding certainty. Afterward, measure, learn, and be willing to reverse.

• Idea meritocracy: Make reasoning inspectable. Weight input by demonstrated competence, not rank. Reward people for surfacing better ideas—even when it stings. This builds trust and improves hit rate over time.

Both models institutionalize a simple ethic: ego last, evidence first.

How to spot them in a meeting

• They change their mind in real time when presented with better data—and tell you exactly why.

• They tell clear “earned secrets” from customer trenches, not abstract market takes.

• They narrate past failures as upgraded beliefs, not blamed circumstances.

• They move effortlessly between 10-year vision and this-quarter KPI mechanics.

• They ask for the intro they need tomorrow and already have a plan if it doesn’t land.

A founder’s practice plan

• Install a kill-switch: predefine metrics that trigger a pivot or sunsetting.

• Run red-team reviews: schedule a monthly “why we’re wrong” session led by a dissenter.

• Track decision memos: hypothesis, evidence, decision, outcome, lesson. Close the loop.

• Ban absolute language in analysis. Replace certainty with probability.

• Make “I was wrong” a badge. Reward it publicly.

The punchline

I've spent years pattern-matching across startups, digging through founder trajectories, and watching ecosystems evolve. But nothing crystallizes the proximity advantage quite like watching the current AI wave unfold in San Francisco. If you're an AI founder operating outside the Bay Area right now, I'll cut to the chase: you're likely working with information that's 6 to 12 months behind what the top practitioners already know. That's not speculation—it's a measurable information lag that shows up in research adoption patterns, model access timelines, and the velocity of knowledge transfer through dense networks.

Let me explain why this matters and what you can do about it.

The Psychology Isn't Vibes—It's Data

The "you become who you surround yourself with" principle isn't motivational poster material—it's backed by serious research. Studies tracking thousands of people show that simply sitting next to someone increases friendship probability from 15% to 22%. Harvard psychologist David McClelland put a number on it: the people you habitually associate with determine as much as 95% of your success or failure.

In tech, this compounds fast. Behaviors spread through networks like viruses—when everyone around you is raising big rounds and thinking in 10x terms, that recalibrates your entire operating system. When your coffee-shop neighbor just closed a Series A and your gym buddy is scaling to 100 engineers, mediocrity stops being an option.

The 12-Month Knowledge Lag: Not a Theory, a Pattern

Here's where it gets concrete. Analysis of AI research publications between 2000-2010 revealed that China's research topics systematically lagged the U.S. by several years. Despite massive investment and eventually matching publication volume, China's choice of research topics more closely resembled what the U.S. was working on in previous years than the current year.

This isn't about capability—it's about information flow architecture. The U.S., and specifically the Bay Area, sets the agenda. Everyone else follows with delay.

Now layer on the insider advantage. OpenAI and Anthropic provide early access to new models for select groups—sometimes 6 to 12 months before public release. Recent reports indicate OpenAI employees have been testing GPT-5 capabilities internally while the rest of the world is still optimizing for GPT-4. Anthropic runs similar beta programs with hand-picked customers—GitLab, Midjourney, Menlo Ventures—who get to build on capabilities that won't be widely available for months.

Translation: while you're reading the release notes, insiders already shipped v2 of the thing you're just starting to prototype.

I Watched This Play Out at YC Demo Day

Earlier this year I attended YC's S25 Demo Day at their Dogpatch HQ in San Francisco. One minute, one slide, 150+ companies—most of them AI-native. What struck me wasn't just the quality (though 92% being AI-focused is wild). It was the velocity of information exchange.

Between pitches, I grabbed tacos from the food trucks and ended up in a 10-minute conversation with a founder who casually mentioned they'd been testing an unreleased model variant for two months. Another founder referenced a research technique I wouldn't see published until weeks later. These weren't secrets—they were just the baseline of what's considered "current" when you're embedded in the ecosystem.

That's the gap. It's not dramatic; it's cumulative and silent.

SF Is Where the Information Density Peaks

The numbers back up the anecdote. San Francisco pulled in over $29 billion in AI venture funding in the first half of 2025 alone—more than double the previous year and vastly outpacing every other city globally. Nearly 50% of all Big Tech engineers and 27% of startup engineers live in the Bay Area. OpenAI signed 500,000 square feet of office space and is hunting for more. Anthropic, Pika, Character.AI, and dozens of unicorns operate within walking distance.

This isn't just about talent density—it's about information flow velocity. One AI observer I follow mentioned attending an average of three AI-focused events per week in the Valley. Monthly Silicon Valley meetups on GenAI, LLMs, and agents pack rooms with founders, researchers, and VCs. That's 12+ high-bandwidth information exchanges per month, multiplied across thousands of participants. Lu.ma and X.com are bibles around here.

Knowledge doesn't spread through press releases—it spreads through repeated, high-trust, face-to-face interactions. The famous Allen Curve proves it mathematically: communication frequency drops exponentially with distance. In practice, this means a 30-minute coffee in SF with someone from DeepMind or Anthropic can shift your product roadmap in ways a dozen Zoom calls never will.

The Problem-Solving Velocity Nobody Talks About

Here's something that doesn't show up in funding announcements but matters enormously: the speed at which you can solve technical and operational problems in SF is orders of magnitude faster than anywhere else.

The Bay Area concentrates 35% of all AI engineers in the United States—Seattle, the second-densest hub, has only 23%. But it's not just the raw numbers; it's the depth and diversity of expertise. Your angel investors aren't just capital allocators—many are former CTOs who've debugged distributed systems at scale. Your advisors have shipped ML models in production at Google, Meta, or Anthropic. Your neighbor in the coworking space solved the exact infrastructure bottleneck you're hitting right now.

I've seen this play out repeatedly. A founder hits a gnarly RLHF training issue on Thursday afternoon, texts an advisor who used to run safety at a LLM unicorn, and by Friday morning has three potential solutions plus an intro to someone at Hugging Face who's dealt with the exact edge case. That 18-hour turnaround doesn't exist in other ecosystems—not because the expertise doesn't exist elsewhere, but because the density and accessibility of that expertise is unmatched.

Corporate VCs operating in the Bay now provide active mentorship, market access, and infrastructure resources beyond just checks. When you're stuck on a technical decision—whether to fine-tune versus RAG, how to architect your agent orchestration layer, which inference provider to use—you're not Googling or posting in Discord. You're texting someone who's already made that exact decision at scale and lived with the consequences.

The operational side is equally compressed. Hiring your first head of sales? Your investors can intro you to three candidates by Monday who've scaled GTM at AI companies. Need to navigate SOC 2 compliance? Someone in your YC batch just went through it and will walk you through the checklist over coffee. Fundraising strategy? Your advisor literally closed a $50M Series B last quarter and knows exactly what metrics Sequoia is asking for right now.

This isn't networking—it's operational infrastructure disguised as relationships. And it only works at this velocity when everyone is physically close enough for spontaneous problem-solving.

Real Founders Are Making the Move—And Saying It Out Loud

The migration patterns tell the story. AI founders from Canada, Europe, Asia, and Latin America are relocating to SF in unprecedented numbers. Indian VCs like Elevation Capital and Peak XV are opening SF offices specifically to stay close to AI developments.

Emmanuel Martes moved his fintech startup from Bogotá to San Francisco and captured it perfectly: "Everywhere else, you're a weird person who wants to start a company. Here, everyone is building".

Ben Su, a Canadian entrepreneur building an AI lawyer, explained his move: "We're hitting the ceiling in Canada, and the mecca of the startup world is in San Francisco". While Canada raised less than $5 billion across all startups, the Bay Area alone pulled $27+ billion in AI funding.

The NYT recently profiled the wave of 20-something founders flooding SF—many dropping out of MIT, Georgetown, and Stanford specifically to be in the city during the AI boom. Jaspar Carmichael-Jack moved to SF, built Artisan AI, and scaled it past $35 million in valuation. Brendan Foody left Georgetown at 19, raised millions for Delv, and is now hiring dozens in the Arena district near OpenAI's headquarters. These founders didn't just relocate geographically—they relocated into an operational support system that accelerates everything.

The pattern is clear: ambitious founders are voting with their feet because proximity compresses time.

The Compound Advantages Are Structural

When you're in SF, several things happen simultaneously:

Early Model Access: Companies like OpenAI run early access programs for safety researchers and select partners. If you're local and networked, you're in the room when capabilities get previewed. That's a 6-12 month product development advantage over teams working with publicly available tools.

Conference Intel: Major AI conferences like NeurIPS and ICML create informal knowledge-sharing loops around presentations and workshops. Attendees get advance briefings, hallway demos, and pre-publication insights that never make it into the proceedings. Being there in person means absorbing trends months before they're documented.

Talent Movement Signals: When a key DeepMind researcher joins Anthropic or an OpenAI engineer spins out a new company, the implications are immediately obvious to insiders. You hear about pivots, technical breakthroughs, and capability jumps through informal networks before they're announced publicly.

VC Intelligence Networks: Bay Area VCs don't just write checks—they aggregate intelligence across dozens of portfolio companies. When Sequoia or a16z share pattern recognition about emerging trends, they're synthesizing confidential data from hundreds of startups. That intelligence doesn't exist in other ecosystems.

Face-to-Face Conversion Rates: Research shows face-to-face requests are 34 times more successful than email. For fundraising, recruiting, and partnerships, being in the room isn't a nice-to-have—it's the difference between a warm intro and a cold outbound.

Expert Problem-Solving Speed: With 50% of Big Tech engineers concentrated in the Bay Area, the time between "we're stuck" and "here's how to fix it" collapses from weeks to hours. Your advisors and investors aren't just cheerleaders—they're active debugging partners who've already solved your exact problem.

The Cost of Distance

While remote work democratized access to global talent, it also revealed the irreplaceable value of physical proximity. Virtual collaboration tools cannot replicate the spontaneous interactions that drive innovation. The most breakthrough ideas often emerge from unplanned conversations—the coffee shop encounter that becomes a partnership, the hackathon that spawns a unicorn, the demo day that attracts unexpected investors.

77% of employees who work remotely show increased productivity in routine tasks, but innovation requires the serendipity that only physical proximity provides. When groundbreaking AI research is being discussed in San Francisco coffee shops and exclusive invite-only dinners, remote founders miss critical insight and opportunity.

The knowledge lag compounds over time. Research shows that informal knowledge-sharing mechanisms—professional networking events, mentorship programs, and casual interactions—are critical drivers of AI innovation. When these interactions are geographically concentrated, outsiders operate with systematically outdated information that affects fundamental business decisions.

More critically, when you hit a technical wall at 11pm and need someone who's debugged transformer architecture issues at production scale, the difference between texting an advisor two blocks away versus posting in a Slack channel with global time zones can mean the difference between shipping Monday or next month.

What This Means If You're Building

The playbook isn't complicated, but it requires commitment:

Establish a Physical Presence: You don't need to move your entire team overnight, but having founders and key decision-makers in SF for sustained periods is non-negotiable. Aim for 6-12 week sprints aligned to model release cycles, major conferences, and fundraising windows.

Default to In-Person for High-Stakes Interactions: Investor pitches, lighthouse customer meetings, senior IC recruiting—do these face-to-face whenever possible. The conversion delta compounds over quarters.

Build a Local Advisory Lattice of Technical Operators: Surround yourself with practitioners who are one hop from frontier labs, leading research groups, or policy/safety desks. Prioritize advisors and angels who've actually built and scaled AI systems in production—their ability to help you debug architectural decisions or navigate technical tradeoffs in real-time is worth more than their capital.

Prioritize Networks Over Newsfeeds: The most valuable information never hits TechCrunch. It spreads through meetups, invite-only dinners, hackathons, and coffee chats. Treat your calendar like an operating system—weekly office hours with VCs, monthly customer deep-dives, quarterly recalibrations based on new model capabilities.

Weaponize Geographic Proximity for Speed: When you're blocked on a technical decision, use the density advantage. Text an advisor, grab coffee with a portfolio founder who's been there, or walk into an investor's office with your laptop open. The 18-hour problem-solving loop only exists in SF.

Bottom Line

The proximity principle that governs friendships also governs information access and problem-solving velocity, and in AI, both timing and execution speed are everything. San Francisco remains the highest-signal, highest-leverage surface area in the world for AI—not because of weather or culture, but because knowledge flows 6-12 months ahead of everywhere else and technical problem-solving happens at 10x speed.

If you're serious about building a category-defining AI company, the math is simple: surround yourself with the best, plug into the densest information networks, compress the feedback loop between idea and execution, and tap into the collective technical expertise that can unblock you in hours instead of weeks. That happens in one place right now, and it's not over Zoom.

The future belongs to founders bold enough to position themselves at its center. For AI in 2025, that center is unquestionably San Francisco—where proximity to greatness, exclusive access to tomorrow's breakthroughs, and instant access to world-class problem-solvers becomes the catalyst for extraordinary achievement.

The AI sector in 2025 exhibits classic bubble characteristics, but unlike previous tech manias, this one sits atop genuine technological transformation. Here's what the latest data reveals about timing, risks, and strategic positioning.

The Bubble Evidence Is Overwhelming

Multiple indicators confirm we're in speculative territory. AI startups now trade at 50-70x revenue multiples, while the sector captures 50% of all VC dollars—mirroring dot-com peak concentration. Most telling: AI companies spent $50 billion on Nvidia chips but generated only $3 billion in revenue in 2023, creating a staggering 17:1 investment-to-revenue ratio.

Even industry leaders acknowledge the excess. OpenAI's Sam Altman explicitly warned that "investors are overexcited about AI," while Oracle's recent $10+ billion bond issuance to fund AI infrastructure exemplifies the arms race dynamics driving unsustainable spending.

But This Bubble Has Real Foundations

Unlike the dot-com era's purely speculative companies, today's AI leaders generate substantial cash flow. Enterprise adoption jumped from 55% to 78% between 2023-2024, with measurable productivity gains: 25% speed increases and 40% quality improvements across knowledge work.

The $7 trillion projected investment in data centers through 2030 reflects genuine infrastructure needs, not speculation. Leading AI companies like Nvidia reported 70% year-over-year growth from real customer demand, creating profitable revenue streams that didn't exist during previous bubbles.

Strategic Navigation: The Three-Horizon Approach

Horizon 1 (0-18 months): Quality Focus
VCs should implement defensive positioning—backing only AI companies with strong ARR growth, GTM strategy and proven unit economics. The sweet spot is vertical AI applications capturing 80% of traditional SaaS annual contract values while avoiding foundation model companies requiring $100+ million in compute costs.

Horizon 2 (18-36 months): Consolidation Plays
Focus on application-layer dominance rather than infrastructure. Prioritise companies with defensible moats—proprietary data, network effects, or deep integration lock-in. Prepare for the consolidation wave as weaker competitors fail, creating acquisition opportunities.

Horizon 3 (3-5 years): Next-Wave Technologies
Position for agentic AI, physical robotics, and edge computing as the market matures beyond current foundation model limitations.

Timing the Inevitable Correction

Historical patterns suggest correction timing within 6-18 months of peak warning signs. Given current indicator convergence—extreme valuations, performance gaps, and industry warnings—expect significant turbulence beginning late 2025 through mid-2026.

However, this correction may be more selective than previous crashes. Companies with real revenue and proven business models could weather the storm, while speculative players face 50-80% value destruction.

The Bottom Line

The AI bubble is real, but its resolution will likely be messier and more uneven than clean historical parallels. Smart capital should prepare for significant market turbulence while recognizing that underlying AI transformation remains genuine—current valuations and expectations just need dramatic recalibration.

I've been knee-deep in the startup world for years—founding companies, advising founders, crunching data science models to spot patterns before they become trends. But nothing quite prepared me for my first Y Combinator Demo Day in 2025. Held at their sleek new offices in San Francisco's Dogpatch neighborhood, it was a masterclass in production value: seamless tech, high energy, and an atmosphere that screamed "this is where the future gets built." If you're wondering why YC continues to dominate, let me break it down from an insider's view—complete with the chaos, the connections, and the signals that have me more excited than I've been in years.

The Setup: More Than Just Pitches

Picture this: a full day of back-to-back presentations, each founder getting exactly one minute and one slide to pitch their "life's" work. Breaks were strategically timed for networking, with food trucks parked for breakfast and lunch—think gourmet tacos and craft coffee fuelling deal talks under the California sun. The energy was electric, a mix of nervous excitement from founders and calculated intensity from investors. YC's Dogpatch HQ felt like a tech temple: modern, spacious, and designed for serendipitous collisions. No wonder they pulled off a production this polished; it wasn't just an event, it was an ecosystem accelerator on steroids.

Shoutout to the unsung heroes—the technical support team behind the Demo Day investor portal. When a few of us hit demo day account access glitches (hey, even the best tech has hiccups), they resolved them in real-time, mid-event. That's the kind of operational excellence that keeps the machine running smoothly.

The Founders: Youth, Tech, and Raw Ambition

What struck me most were the founders themselves. This batch skewed young—seriously young. I'd estimate 15-20% were college dropouts or fresh grads who'd already racked up impressive feats since high school: building apps that scaled to millions, hacking together AI prototypes in dorm rooms, publishing AI research papers or launching side hustles that caught YC's eye. At most, I spotted three with MBAs; the rest were deeply technical, often with backgrounds in engineering, CS, or data science. These aren't polished executives—they're builders who code first and pitch second.

Take the companies from the S25 graduates list I reviewed: teams like BootLoop (firmware in minutes via AI) or Janet AI (an AI-native Jira alternative) exemplify this. Founders aren't waiting for permission; they're leveraging AI to solve real problems in dev tools and beyond. It's refreshing—no corporate fluff, just raw innovation from people who grew up with GitHub as their playground.

One subtle tell? Look at their company URLs. Domains like phases.ai, getlilac.com, or bootloop.ai aren't flashy .coms or clever wordplays—they're straightforward, often incorporating "AI" directly. It screams product focus over marketing polish. These founders prioritize building something that works over snagging the perfect brand name or premium URL. In an era where AI lets you prototype in days, that no-nonsense approach is a competitive edge.

The Room: Celebrities, Athletes, and Serious Money

The investor crowd was a who's who of tech and beyond. I spotted comedian Hannibal Buress scanning pitches with a notebook in hand, and Ty Montgomery—the former NFL football player and Stanford alum—deep in conversation about deals. It wasn't just VCs; it was a melting pot of cultural influencers, athletes, and operators who see startups as the next big bet. That diversity amps up the energy—suddenly, a quick chat over lunch could lead to a celebrity endorsement or a strategic partnership.

For me, the real magic was in the outreach. Out of the 150+ pitches, I reached out to 20-30% to express interest in chatting further. That's by far my highest engagement rate in years. Why? The format forces clarity: one minute, one slide strips away the noise, letting the core idea shine. Many pitches hooked me instantly—they're not just building products; they're redefining categories with AI at the core.

Why YC Is Skyrocketing—and What It Means for the Rest of Us

Personally, I think YC is about to pull even further ahead. Their network is unmatched: a global web of alumni, investors, and experts that feels like they're living in the future. What I do with machine learning—spotting patterns in markets, predicting startup successes—they're embodying it daily through relentless iteration and founder support. This batch's AI dominance (90% of pitches were AI-centric, per reports) isn't hype; it's a signal. We're seeing agentic systems, voice AI, and edge computing solve real problems in healthcare, fintech, and dev tools.

Three quick provocations for anyone in the startup game:

Embrace the Youth Wave: If 20% of top founders are fresh out of school, rethink your hiring. Technical depth plus unjaded ambition is the new superpower.

Network Like It's Lunch: Events like this prove serendipity scales. Skip the formal meetings; real deals happen over food trucks.

Bet on Clarity: In a world of infinite prototypes, the winners distill complexity into one slide. That's the AI-era moat.

Demo Day Summer 2025 wasn't just an event—it was a glimpse into tomorrow. If this is YC's trajectory, count me in for the ride. 

Quick Summary:

Executive Overview:

🤖 AI Dominance: Unprecedented Concentration

Key Finding: 92.2% of companies are AI-focused

• 127 companies (83.0%) explicitly in "Machine Learning / AI" category

• 141 companies (92.2%) mention AI in descriptions or categories

• This is the highest AI concentration in YC history

Top AI Use Cases:

1. AI Agents - 30 companies (21.3%) - Autonomous software workers

1. AI Automation - 12 companies (8.5%) - Workflow automation

1. Voice AI - 9 companies (6.4%) - Voice assistants

1. AI Analytics - 9 companies (6.4%) - Data analysis tools

🏢 Business Model Analysis

B2B Enterprise Focus: 74.5%

• 114 companies target business customers

• 84 companies (54.9%) are SaaS businesses

• 150 companies (98.0%) target global US/World markets

🎯 Key Trends Identified

1. The AI Agent Revolution

• 30 companies building autonomous AI workers

• Replacing human tasks across industries

• From customer service to software development

2. Developer Tools Renaissance

• 33 companies (21.6%) building dev tools

• "Cursor for X" pattern emerging

• AI-powered coding assistance everywhere

3. Vertical AI Solutions

• Healthcare: AI EMRs, clinical trials, pharmacy

• Finance: AI accounting, billing, analysis

• Real Estate: AI property management

• Manufacturing: AI operations

4. Enterprise Software Transformation

• Traditional tools rebuilt AI-native

• Focus on knowledge work automation

• Workflow automation emphasis

A fresh conversation on the No Priors podcast has Andrew Ng declaring product management the new bottleneck in AI startups. His core thesis: while engineering teams can now build prototypes "10x faster" thanks to AI-assisted coding, product teams are still operating with pre-AI workflows, creating a dangerous mismatch that's "really painful" for founders.

Lately, I've been fielding questions about venture studios – those ambitious setups that aim to manufacture startups at scale. On paper, they look revolutionary: higher success rates, juicier IRRs, and a systematic approach to innovation. But as someone who's evaluated countless models for our portfolio, I can tell you the reality often falls short, especially when viewing them as an asset class for serious fund returns. Today, let's unpack the evidence why most don't work, and the rare formula that makes some succeed. This isn't armchair theory; it's grounded in data, case studies, and insights from my own deal flow.

The Seductive Promise – And Why It Often Crumbles

Venture studios burst onto the scene promising to fix VC's biggest pains: inconsistent dealflow, high failure rates, and inefficient capital deployment. The stats are eye-catching – studio-backed companies reach Series A 72% of the time versus 42% for traditional startups, with IRRs hitting 53% compared to 21.3% norms. They shave timelines too, getting to Series A in about 25 months instead of 56. I've seen pitches where studios position themselves as "startup factories," and honestly, who wouldn't want that in their portfolio?

But here's the rub: most flame out within 24 months. That's not random bad luck; it's baked into the model. In my experience reviewing studio proposals, the hype masks deep structural flaws that make them a risky bet for LPs seeking scalable, reliable returns. Let's break it down with the evidence.

The Structural Cracks That Sink Most Studios

From stakeholder wars to capital headaches, these issues aren't edge cases – they're systemic.

The Impossible Balancing Act of Four Stakeholders

Traditional VCs have it simple: LPs provide capital, GPs pick winners, founders execute. Studios? They must satisfy entrepreneurs (who want autonomy and upside), studio staff (needing comp and growth), follow-on investors (demanding clean terms), and LPs (chasing returns). It's a recipe for conflict. I've passed on studios where entrepreneurs felt like cogs in a machine, leading to talent flight and diluted innovation. Data shows this tension torpedoes governance and decision-making.

Resource Dilution: Great in Theory, Impossible at Scale

Studios tout "hands-on support" as their secret sauce, but the numbers don't add up. Launch one company a week with 100 staff? That's barely two full-timers per venture – hardly the deep operational help promised. Scale up, and quality tanks; stay small, and you can't deploy enough capital for fund-level returns. I've seen this play out in SEA studios that overextend, ending up as glorified consultancies rather than return generators.

Fundraising Nightmares and Wonky Capital Structures

Evidence from 2020-2024 reveals traditional VCs are 1.6x more likely to close funds than studios. Why? Complex valuations create the "valuation trap," higher fees scare LPs, and perceptions of "dead equity" deter later rounds. Add longer timelines and regulatory ambiguities (fund or operating company?), and you're burning cash before deploying it. In my due diligence, this often signals execution risk – markets shift, and the studio's left holding the bag.

Why Institutions Steer Clear

For big allocators like pensions or endowments, studios are a tough fit. Most cap at sub-$200M funds, too small for meaningful commitments, and their sector-specific focus creates concentration risks. Metrics suffer from survivorship bias too – we only hear about winners, not the ideation failures. Corporate studios? Even worse, often becoming "innovation theater" bogged down by parent company bureaucracy and mismatched incentives.

The Winning Formula: What It Takes to Make Studios Work

Not every studio crashes and burns. From my vantage point, the successes – think specialized operators in niches like fintech or healthtech – share a rigorous playbook. Here's the evidence-backed essentials, scored by their impact based on research and my observations:

• Strategic Focus & Specialization (9.2/10): Ditch the generalist approach; vertical-agnostic studios succeed only 19% of the time. Winners leverage proprietary edges like data access or industry networks – crucial in emerging markets like SEA.

• Operational Excellence & Proven Playbooks (9.0/10): Codified processes for everything from ideation to scaling. This includes stage-gates that validate ideas early, reducing waste and accelerating paths to revenue.

• Significant Ownership Stakes (8.8/10): Holding 30-50% equity justifies the ops investment and captures value from solid exits, not just unicorns.

• Experienced Team & Leadership (8.7/10): A mix of serial entrepreneurs, functional experts, and investors. Weak teams are a red flag in my evals – you need proven builders to navigate the chaos.

• Quality Control Systems (8.6/10): Strict validation thresholds for market fit, tech viability, and business models. This kills duds fast and preserves capital.

• Proper Governance & Alignment (8.5/10): Dual structures separating ops from investing, with clear equity rules to align all stakeholders.

• Adequate Capitalization (8.3/10): Patient, reserved funding – balance sheet style, not skimping on ops budgets.

• Market-First Approach (7.9/10): Prioritize customer validation over building; it's the best defense against vanity projects.

Implement all eight? You've got a shot at outsized returns. Skimp on any, and you're statistically doomed.

Wrapping Up: A VC's Practical Advice

Venture studios aren't the VC extinction event some predict – they're a niche tool, better for hands-on operators than broad institutional plays. The evidence shows their flaws often outweigh the upsides for fund returners, but in the right hands, they can be game-changers, especially in underserved regions.

If you're building or backing one, stress-test against these factors. 

The SGX Listing Drought Is Real

Tech IPOs on SGX are unlikely to surge without major fixes. In 2024, just four listings occurred, none on the Mainboard, raising only $31 million2. By mid-2025, only three tech-related listings have materialized amid a global IPO rebound3. Rigid rules, such as the S$30 million profit threshold, exclude cash-burning tech firms focused on growth.

SGX's daily trading volume stalls at $1.1 billion4, contrasting Singapore's 7th global startup ranking and $144 billion in value1. This disparity drives firms like Grab ($40 billion NASDAQ debut) and Sea Limited (NYSE) abroad. Conservative investors prioritize dividends, with 68% of trades from volatility-averse retail players5. Retail outflows, like S$189.9 million in late 2019, exacerbate thin liquidity6.

Regional Variations: Not Just a Singapore Slump

Asia-Pacific peers outperform. Tokyo Stock Exchange (TSE) adapts swiftly, with $273 billion in Growth Market volume in 2023 extending into 20257. Its Asia Startup Hub aids 14 regional firms via streamlined processes SGX lacks.

Jakarta's Indonesia Stock Exchange (IDX) booms: 17 tech IPOs in 2023 climbed to 22 by mid-2025, with $881 billion market cap8. GoTo's $1.1 billion raise exemplifies flexible rules like dual-class shares. Australia's ASX supports 140 tech firms, enabling Afterpay's rise from $165 million to $29 billion acquisition9.

Geography influences: TSE yields 50% post-IPO gains, IDX's retail drive boosts 59% volume, and ASX's principles-based governance reduces bureaucracy10. SGX operates at 75% below tech potential (±7% confidence, DealStreetAsia and McKinsey data)11.

Snapshot:

The Core Barriers: Rigidity and Risk Aversion

Outdated profit mandates ignore models like Amazon's, deterring tech startups12. Low liquidity repels institutions12, while risk aversion clashes with tech's experimentation13. Investor skew toward dividends undervalues tech, with 86% of Catalist stocks below debut price due to limited institutional buy-in14.

Funding shortages for Catalist firms (median revenue ~S$27.4 million) create stagnation, as banks favor traditional sectors with scant tech coverage versus Hong Kong's robust analysis15. Early-stage gaps, like Temasek's 88% investment cut from 2021-2024, favor US exits16. Regulatory delays (4-6 months under MAS/SGX RegCo) lag rivals, worsened by global pressures like high rates and no new unicorns in 202317.

Reforms emerge: SGX's profit cut to S$10-12 million and dual-class shares lag TSE/IDX innovations, but McKinsey eyes 150% regional listing growth by 2027 with metric shifts like revenue focus—though volatility risks persist (15% ASX dips)10.

Standout Successes That Defy the Odds

Successes highlight potential. TSE's JDRs let Singapore's Omni-Plus System list seamlessly18. IDX's Bukalapak raised $1.5 billion via flexible exits19. ASX's WiseTech Global scaled globally20.

These (20% of regional IPOs)10 prove innovative regulation works, contrasting SGX's rigidity. IDX's paths offer scaling lessons amid Singapore's talent and cost hurdles2.

Wisdom from Market Makers

Leaders urge evolution:

• Temasek's Rohit Sipahimalani: "SGX must adapt to capture tech value or lose out"3.

• TSE: "Flexibility drives 80% IPO share"7.

• DealStreetAsia: "IDX's 22 IPOs show retail power—SGX needs it"5.

• ASX: "Principles-based rules attract 230 listings"9.

What This Means for Singapore

Without change, SGX misses 70% of Southeast Asia's $300 billion startup value by 203012. Reforms like tech boards, Catalist funds via MAS's S$5 billion program, and coverage boosts could target <20% retail dominance and >$2 billion volume11. Startups: Avoid SGX's liquidity pitfalls; favor TSE/IDX gains. Advances could empower regional unicorns, addressing talent via incentives2.

Revitalizing Southeast Asia's Startup Ecosystem: From Ideation to Deep Tech

To fully address SGX's shortcomings, the broader regional startup ecosystem must be revitalized, tackling gaps from ideation to deep tech R&D funding, spinouts, and venture capital (VC) performance issues. Southeast Asia faces funding shortages for early-stage startups, talent constraints in AI and data science, and regulatory fragmentation across jurisdictions12. Deep tech funding tumbled 34% in 2024, yet its share of VC activity rose to 17.6%, driven by health tech and biotech, though challenges like skilled personnel shortages and high development costs persist717.

Spinouts from research institutions struggle with insufficient design data, manufacturing delays, and market penetration in areas like Singapore and Vietnam18. VC firms exhibit lackluster performance, with equity investments down amid selectivity for sustainable models over aggressive growth56. Quality issues include poor due diligence, fraud risks (e.g., eFishery case), and a shift to capital efficiency imperatives916.

Key actions include: Boosting ideation through university partnerships and internal talent development2; Increasing deep tech R&D funding via government incentives and green tech funds411; Facilitating spinouts with standardized governance frameworks and cross-border enforcement9; Addressing VC quality by embedding sustainability, enhancing transparency, and diversifying revenue streams25. Initiatives like ASEAN's Digital Economy Framework Agreement could finalize in 2025 to enable greater collaboration5. These steps could accelerate innovation, with projected GDP growth of 4.7% supporting consumer spending and ecosystem resilience5.

The Path Forward: Critical Reforms Needed

Turning SGX around requires aggressive action—it's possible but demands commitment amid economic headwinds. Top priorities: Flexible regulations (disclosure-based shift), liquidity boosts (S$5 billion fund, tax rebates), and ecosystem enhancements (research, talent support, VC governance)91114.

Horizon shifts:

• Reform Wave: Dual-class expansions could double listings by 202711.

• Regional Alignment: Mirror ASX's institutional mix via IDX pacts8.

• Innovation Edge: TSE-like hubs halve times to 12 months12.

2026-2030 outlook, cautiously:

• Tech Surge: Capture 30% unicorns, adding $50 billion cap11.

• Liquidity Leap: $3 billion volumes matching ASX4.

• Global Ties: Pacts boost 40% non-local listings10.

• Risk Tools: AI cuts volatility 25%12.

The revolution spreads—Singapore can lead with adaptive exchanges and a robust ecosystem. What reforms do you prioritize? What else is needed?

This draws from 2024-2025 reports by Bloomberg, DealStreetAsia, McKinsey, PitchBook, and exchange filings. (Word count: 1,128)

The Global IT Spending Landscape: Size and Scale