Jeff (Startup Whisperer)

For years I have said Singapore deep tech underperforms because our research is not world class. So I went looking for evidence to prove me wrong, and it half did. Our science passes the academic test. It fails the investor's test: does anyone pay for it?

In the Leiden Ranking 2025, 18.2 percent of NTU's papers and 17.1 percent of NUS's are in the world's top 10 percent most cited. Oxford is at 18.2, the University of Tokyo at 8.2. We publish like Oxford, not Japan. The scientists are not the problem.

Now the money. Public-sector licensing revenue was S$18.6 million in 2023, down from S$24.2 million, which the RIE2025 report called "significantly below" MIT, Harvard and Stanford. By its own figures, that is about 0.4 percent of public R&D. UK universities earn 3.0 percent of research spending from IP, 2.4 percent without spinout share sales, and US universities 3.4 percent. Our entire public research system earns less from licensing than Stanford, at over US$88 million in fiscal 2025, does alone. So the break sits between the paper and the buyer.

The exits agree. The largest Singapore public-research spinout acquisition I can find is A*STAR's Advanced Micro Foundry, bought by GlobalFoundries for US$453 million. In 2025, two Oxford spinouts sold for over US$1 billion: OrganOx at US$1.5 billion and Oxford Ionics at US$1.075 billion. A*STAR spinout Mirxes, which the RIE2030 factsheet says was valued above US$1 billion, has been suspended in Hong Kong since April 1 at the company's request, while it works through exchange conditions that include an independent forensic investigation into prepayments and material cash outflows, which as of October 1 was still ongoing under new advisers.

I think the break comes before the research starts. Only 7.2 percent of NUS papers and 7.3 percent of NTU's are co-authored with industry. With SUTD, they are the bottom three of the eleven universities I compared. MIT is at 13.9. Some corporate-lab work stays unpublished under confidentiality, but its valuable IP should still show up in licensing income, which is falling. Without a customer in the room, you get excellent papers nobody asked to buy.

Lucence, an A*STAR liquid biopsy spin-off, put the buyer first. With government support, it set up its US headquarters and lab in Palo Alto. By the government's account, that let it win US Medicare coverage for its flagship test in 2023, the first Asian healthcare services provider to do so, and revenue doubled to over US$6 million within a year. It is not a unicorn. But its science was validated by a foreign payer deciding to pay, not by a citation count.

Our scoreboard does not count that. Asked in Parliament what share of deep tech projects become commercially viable, Minister Gan Kim Yong said "There is no single or simple metric of commercial viability." A*STAR's KPI table counts spin-offs, not what they become. We measure the start, so we fund the start.

Even the starts are thin. NUS's GRIP gives up to S$250,000 per startup, and its 121 companies since 2018 have secured nearly S$70 million in external funding and grants. Oxford's spinouts raised £489.8 million in one year. People are scarcer still. Entrepreneur First left in 2022, deciding "Singapore does not have the supply of talent" for several programmes at once. In my experience, so is deep tech pattern recognition among investors here, mine included.

Now the objections. Licensing lags research by ten to fifteen years, yet Singapore has committed more than S$125 billion to research plans since 1991, including RIE2030, and income still fell. The Minister says corporate labs pay back by training researchers and anchoring research hubs here. True, but that is the subsidy's return, not the IP's. Asked what share of that IP earns zero or nominal licensing, the answer was that about 80 percent of projects have licensing revenue "expected." Nor can outsiders check: NUS last published licensing income for FY2017, and A*STAR's latest annual report gives none.

NUS Enterprise became NUSX on September 24, with Nova, an AI licensing engine, and a 2035 target of one venture with US$100 million in revenue or 1,000 jobs. It is the first research institution here I know of to publish value targets, which deserves credit. But one such company in nine years is a low bar, there are no interim milestones or budget, and Nova chases the old KPI. Its "10 unicorns" come largely from student and BLOCK71 programmes. At a few weeks old, judge it by its own scoreboard.

MP Kenneth Tiong proposes that publicly funded IP not be "permanently transferred offshore without prior approval." I share his diagnosis, not his fix. Lucence found its proof in Palo Alto. Validate abroad, and anchor here through equity and headquarters, not IP lock-in.

Two fixes matter most. Grants should require a foreign customer, pilot or lead investor within 12 to 18 months, with growth money rewarding a Singapore holding company and headquarters. And NRF should publish venture value for every RIE-funded spinout yearly, as Oxford does for its own, while research institutions resume publishing licensing income. Beyond those: standard spinout terms like the UK's USIT guide, operator-led venture studios, and company creation counted in academic careers, as A*STAR proposes. Investors like me should underwrite customers, not citations.

Our science is ready. Our customers are not yet in the lab. Put them there. I believe there is still some work to do but by working together we will get there.

In October 2023, four MIT students who had graduated the year before raised an $8M seed for a code editor. The OpenAI Startup Fund led the round and Nat Friedman came in alongside it. The founders had come through an OpenAI accelerator, so the people who wrote the first check had already watched them work. It was a seed priced like a seed, for a product competing with tools every developer already had. Two years later the company behind Cursor raised a $2.3B Series D at a $29.3B post. The investors who made money on that were not the ones who chased the Series D. They were the ones in the room when four young engineers asked for $8M.

I keep coming back to that round because it is the opposite of what most new investors in the United States think AI investing looks like. They read that OpenAI raised $122B in March 2026 and assume the game is about getting into the big round. It is not. The headline number is the least useful number in the market.

So I looked at the beginning instead. I took 38 American AI companies founded in 2020 or later that have since been valued at $1B or more, and asked who the founders were, how big the first round was, and who wrote the first checks. The list was built by hand and leans toward survivors, so it is not a census. It is enough to see the shape.

The shape is two games that barely touch. The first is the lab spinout. Thinking Machines Lab opened with a $2B seed. Safe Superintelligence opened with $1B. Not one of the lab spinouts I could trace started under $10M. Those rounds are assembled before they are announced, by firms that already know the researchers.

The second game is the one that matters. Fifteen of these companies started with a first round under $10M. Perplexity, Mercor, Harvey, Decagon, Modal, Distyl and Cursor all began that way. And when I looked at who those founders were, it was not who I expected. Most were not dorm room prodigies and not famous researchers. They were engineers and operators out of big companies: Palantir alumni at Distyl, a former Spotify engineer at Modal, Meta and Google alumni elsewhere. People who already knew how to ship and had not yet been priced as stars. That is the gap a new fund can live in.

Then the question is who is already standing in that gap. Here I distrust the usual answer. Count every first round in the set and a16z looks dominant and Sequoia looks strong. But volume flatters whoever writes the biggest checks. Cut it to the rounds under $10M and the picture changes. The OpenAI Startup Fund was in three of them. Nat Friedman and Daniel Gross were in three. a16z was in three, out of twelve first rounds overall. Sequoia was in none; its smallest first check in this set was $25M. The people worth standing next to are the ones who keep showing up when the round is small and the founder is still unknown, and that is a short list.

So here is what I would do if I were starting a new fund in the US today.

I would pick a lane and stay in it: technical operators leaving big companies, young technical founders, and repeat application founders, all raising seeds under $10M. I would write checks of $250K to $1M and be the useful second check a lead actually wants, not pretend to out brand a16z. I would be in San Francisco, because the Bay Area took $122B of the $159B of US AI funding in 2025, and you cannot see deals first from a distance when the concentration is that extreme. I would stay close to the feeder programs and operator networks, because Cursor came through an OpenAI accelerator, Mercor through the Thiel Fellowship, and Distyl out of Palantir. The deal usually starts years before the round. And I would track early signals with software, because a person reading the news is always late.

None of that is an edge, though. It is table stakes. A new fund needs a reason for a founder to pick its check over a dozen others, and the only reason that works is something the founder cannot get elsewhere.

From where I sit in Singapore, that reason is Asia. Think about what happens to a company like Cursor or Harvey a year after the seed. The product works, American customers are paying, and the next question from the board is where the second market comes from. Asia is large, but it is not one market. Buyers in Japan, Korea, Indonesia and Singapore procure differently, regulators ask different questions about data, and the large enterprises that would anchor a regional business mostly buy through relationships rather than inbound demos. A small team in San Francisco cannot learn that from a deck. An investor who can put the founder in front of the right three enterprise buyers, explain how a regional bank will run procurement, and help hire the first person on the ground is offering something concrete. It does not replace a great lead, but it is exactly the kind of help a seed stage founder remembers when the next round gets tight and allocation is being decided.

The big rounds are not wrong. They are just the wrong place for a new fund to look. Across these winners, the lab spinouts started big and closed, and the rest mostly started small and open, founded by people who had shipped before and were not yet priced as stars. Cursor was one $8M round with the right two investors in it. A new fund wins the same way those investors did: small checks into those founders, next to the handful of people who keep seeing them first, carrying an edge a founder will still value two rounds later.

Singapore wants to be the place startups go global from. Enterprise Singapore pitches Startup SG as "your launchpad to the region and beyond," and Budget 2026 set aside S$1 billion to top up Startup SG Equity for Singapore-based deep tech startups, now including at the growth stage. I share the objective. I think most founders reach for it in the wrong order. If you can start in the US, start there, and then build your Asia-Pacific headquarters in Singapore. That gets Singapore what it wants more often, because the power law does not work well for companies that begin here.

Venture returns come from a handful of companies that get very large, and how large a company can get is shaped early by where it starts. Your first customers decide what you learn and how fast. Your first prices set what every later customer expects to pay. A company that starts in a city-state spends its formative years learning from a small set of buyers and pricing for them, then expands into Southeast Asia, which is several markets with their own rules, languages and price points. By the time it reaches the US, if it ever does, its product and its price have been shaped by the wrong customers.

The capital stack points the same way. In 2025, about $274 billion of startup funding went to US-based companies, 64 percent of the global total by Crunchbase's count. All of Southeast Asia raised $5.4 billion across 461 deals by DealStreetAsia's tally, and Singapore took 91 percent of it. The big growth checks and the acquirers who pay up sit on the American side. A Singapore company with the same numbers as a US peer gets priced as a regional story, and that discount compounds every round.

AI widens the gap. The labs, the design partners willing to run something half-built, and the engineers who have shipped on frontier models are concentrated in the Bay Area, and an AI company lives on iteration speed. Being an ocean away from your best customers is a tax you pay every week. Fortune reported in July on a founder building an AI company in Singapore who moved to the Bay Area after noticing his early customers were already based in, or selling to, the US. The power law made the decision for him, late.

Now look at what happens when a company that has already won abroad comes to Asia. Carta was primarily US-focused before it chose Singapore for its first international expansion and regional headquarters, and that team has grown from about fifteen people to over sixty. Ambiq, a US-headquartered chipmaker, runs its regional R&D and commercial hub here and opened a new Singapore office in February, after its July 2025 IPO. Senior jobs, regional revenue, a global brand, with Singapore running Asia. I would argue it achieves "go global from Singapore" better than ten Singapore-headquartered companies that stall in the region.

The state's own capital already behaves this way. EDBI, which now sits under the state's SG Growth Capital platform, says it invests in "high-growth global technology companies and industry leaders looking to scale in Singapore and across Asia." Its write-up on Ambiq, which it backed in a 2018 Series E, describes "its mandate to anchor global innovators." EDBI is not betting that Singapore must be the birthplace, only the anchor. The startup end of policy should take the hint.

There are real exceptions. If your market is Southeast Asia, start here. Grab, which began in Kuala Lumpur, and Sea built for the region, and local knowledge was the moat. Runway is cheaper here, US visas are a hard constraint for many founders, and Singapore's grants and co-investment are real money. This is not an argument that Singapore is second-rate. It is about sequence, for founders who can get to the US and whose customer is global.

If that is you, incorporate and sell where your first great customer lives, and plan the Singapore headquarters as a milestone you reach with revenue, not a fallback you start from. If you are a GP, underwrite where the first customer and the next round come from, not the founder's passport. If you set policy, count the global companies running Asia from Singapore as wins, and let support follow the founder, so a Singaporean who starts in San Francisco has every reason to bring the regional headquarters home. The region is not a discount to SF. It is not a substitute for it either. Singapore wins at the second office.

The most sophisticated investors on both sides of the Pacific told us this week what they value most right now, and it is not the next model. It is the ability to change their minds. They are not retreating from AI. They are paying to keep the door open.

On October 7, at the Milken Asia Summit in Singapore, Temasek's chief investment officer Rohit Sipahimalani said about half of Temasek's AI exposure sits in public markets, and "we want to increase that number ideally to 70% to 75% so it gives us more flexibility to pivot with changes in the environment, as against being stuck with a private asset which there's little you can do when things are changing." He called an unwinding of the AI trade "the biggest risk," said Temasek does not see it as imminent, and allowed that there could be bumps in 2027. This is not a retreat. In July, Temasek said it aims to more than double AI from 6 percent of its portfolio to up to 15 percent by 2031. What is changing is the mix: more AI, held in a form it can sell.

The same week, the PitchBook-NVCA Venture Monitor showed American LPs making the same trade with different instruments. Established firms, those on fund four or later, took 88.2 percent of US venture capital raised this year, the highest share in PitchBook's data, and PitchBook says that preference "largely reflects liquidity constraints." The number of emerging firms closing a fund fell from 927 in 2022 to 211 through the third quarter. First-time funds raised $4.9 billion across 81 vehicles. The top line looks like a boom: a record $515.8 billion of deal value with AI at 82.7 percent of it. PitchBook's Nizar Tarhuni was blunt about what sits underneath: "for most of the market, the liquidity won't show up." On Forge, shares of companies that last raised in 2021 trade at a median 59 percent discount.

Put these together and the story is not that LPs have soured on AI. They are buying optionality wherever they can find it. Temasek finds it in listed stocks. US LPs find it by recommitting to managers who have already sent cash back, on the theory that a firm that distributed once can do it again. The pattern runs beyond venture, too. In private equity, the most patient institutions are selling to get it: Yale was reported in September to be selling about $1 billion of fund stakes, and HarbourVest was reported on October 1 to be buying roughly $1 billion of fund interests from GIC.

I think this is rational, not panic. When the technology moves this fast, the cost of being locked in rises. A private position you cannot exit is a bet that the world you underwrote will still exist when the position matures, and in AI that world can change inside a single fund cycle. So the premium for holding illiquid private assets has to be higher than it used to be. Most managers cannot show that it is. That is the uncomfortable conclusion inside the 88 percent. Capital is not being allocated by manager quality. It is being allocated by liquidity, and those are not the same ranking.

That should change how everyone on the chain behaves. Founders should drop the assumption that their investors are happy to hold for ten years. Build a cap table that is ready for secondaries, and have a tender-offer plan before someone asks for one, because early holders who cannot get out will eventually act like it. GPs should treat a credible path to returning cash early, through continuation funds or structured secondaries, as part of the product, and the LPs underwriting them should pass on funds whose only exit is an IPO window that, in Tarhuni's telling, keeps slipping further out.

In the US, LPs should be honest that an 88 percent allocation to established managers is a liquidity bet, not alpha selection. Emerging-manager terms are at their best precisely when that cohort is starved, so a disciplined small allocation now is worth protecting. In Southeast Asia the lesson is sharper. Temasek is the allocator this region watches most closely, and if it wants the freedom to pivot, the funds that grew up in its ecosystem should build the exit door into the fund design: secondary programs, explicit DPI targets, continuation options. Family offices committing to regional funds should ask to see that plan before they sign.

Everyone is paying for the exit door. The managers who build one will be the ones who get paid.