Y Combinator’s wishlist vs the Entrepreneurs First talent pool

Hello friends, and welcome to The Unsophisticated Investor! Brought to you by Scott & Rob, the founders of Shuttle!

If you want to invest alongside the VC funds who've backed breakout companies like Revolut, Asana, JustEat, Bolt, Lets Get Checked, Loom, Runna, Charlotte Tilbury, Deel, Aircall, AngelList, Carta, TransferWise and many more, regardless of your knowledge, network or net worth, then sign-up using the link below.

Now, let’s get into it 👇

Y Combinator has just published its Summer 2026 Request for Startups. The opening entry, written by partner Tyler Bosmeny, begins with this:

"Last month, a swarm of cheap Iranian drones took out an AWS data center. Nobody stopped them."

Bosmeny is referring, with some artistic compression, to the strikes of the 1st of March 2026, when Iran's Islamic Revolutionary Guard Corps used Shahed drones to hit two AWS facilities in the United Arab Emirates and one in Bahrain. It was the first publicly confirmed military attack on a hyperscale cloud provider. The damage forced multiple availability zones offline and disrupted services from Snowflake to Emirates NBD to the rideshare platform Careem.

Within weeks, YC put out a request for startups building counter-swarm defence systems. That speed of response, from geopolitical events to a publicly stated investment thesis, is what the Request for Startups is for. It is a map of where Silicon Valley's most influential accelerator thinks capital should be moving, drawn up in something close to real time.

Whether the map describes the territory accurately is a separate question.

What is on the latest list

The Summer 2026 RFS is the most concentrated and AI-dominant edition YC has published. The framing is unambiguous: AI has stopped being a feature and started being the foundation, and the categories are essentially variations on that theme.

Among the fifteen entries: AI-native service companies that replace insurance brokers and accountants rather than provide tools for them. Inference chips designed for the agent loop, where YC notes current GPUs hit only 30 to 40% of utilisation. Counter-swarm defence stacks. Electronics in space, contributed by Starcloud's Philip Johnston. Industrial capabilities on the moon. A "company brain" that turns scattered organisational knowledge into something AI agents can act on. Hardware supply chains capable of matching Shenzhen's iteration speed, which YC describes as orders of magnitude ahead of US equivalents.

The thread is consistent. AI as the operating layer for everything physical, financial, and procedural. Defence as a permanent category. Space as infrastructure rather than spectacle. Service replacement as the next venture wave after software replacement.

Compare this with the 2018 RFS, the previous landmark version. That list had nine new categories: Cellular Agriculture and Clean Meat, Carbon Removal Technologies, Voice Apps, Improving Memory, Longevity and Anti-Aging, Brick and Mortar 2.0, Cleaner Commodities, Supporting Creators, Safeguards Against Fake Video.

The contrast is not subtle.

Eight years on, the scoreboard

Some 2018 categories produced real businesses. Carbon removal generated genuine activity, helped along by Stripe Climate and the Frontier advance market commitment, and several durable carbon removal companies trace back to that period. The 2024 list, which added defence tech as an explicit category, looks prescient: European defence, security and resilience startups went on to raise a record $8.7 billion in 2025, a 55% year-on-year increase.

Other 2018 categories produced very little, despite enormous capital deployment.

Cultivated meat is the most striking example. The category attracted roughly $1.6 billion in venture funding across 2021 and 2022. By 2023, funding had collapsed by close to 80% to under $200 million. New Age Meats failed in 2023. SCiFi Foods ceased operations in mid-2024. Beyond Meat, the public-market poster child for the broader alternative-protein category, lost approximately 99% of its peak share value, falling from over $234 in July 2019 to around 52 cents by October 2025. The map said cellular agriculture was a generational opportunity. The territory turned out to be regulatory bans across multiple US states, supermarket dropouts, and a structural cost problem nobody could engineer past.

Voice apps were another miss. The 2018 thesis assumed a wave of Alexa-first and Google Assistant-first applications, with YC offering portfolio companies direct channels to both teams. The wave did not arrive however. The actual voice opportunity emerged seven years later, after large language models made conversational interfaces work, and it looks nothing like what was on the 2018 list.

Improving Memory and Longevity, both 2018 categories, have produced almost no venture-scale outcomes. The largest longevity bets, including Altos Labs, were not YC companies and were funded outside the accelerator system entirely.

The pattern is consistent. YC's hits are categories where the underlying wave was already forming, and the RFS functioned as recognition rather than prediction. The misses are categories where YC tried to summon a wave that had no underlying physics behind it. Cultivated meat, voice apps, several variants of climate-tech enthusiasm in between.

This is a critical distinction for anyone reading the new list. The RFS is most predictive when it describes momentum already visible in the territory. It is least predictive when it describes the world the partners would like to see, rather than the one being built.

The other model

While YC publishes its map, a different operation has spent the past decade arguing that the map is the wrong artefact entirely.

Entrepreneurs First, founded in 2011 by Matt Clifford and Alice Bentinck, has helped create more than 600 companies valued at over $16 billion collectively, up from $3 billion at its 2022 fundraise. Its own most recent round, closed in March 2026, valued EF itself at $1.3 billion, making the firm a unicorn in its own right. Backers include Reid Hoffman, John and Patrick Collison, Eric Schmidt, Demis Hassabis, Founders Fund, and Greylock.

EF's thesis is the inverse of YC's. There is no annual list of categories the firm wants founders to pursue. Instead the firm recruits exceptional individuals, often before they have a co-founder, an idea, or anything resembling a company, and supports them through the process of finding all three. Talent investing, as Clifford calls it, treats the supply of great founders as the binding constraint, not the supply of great ideas.

The portfolio is what happens when the bottleneck assumption is different. Magdrive, building next-generation satellite propulsion, is backed by Founders Fund. Aztec, the privacy layer for Ethereum, by Andreessen Horowitz. Cleo, the AI financial assistant, by Sofina. Neptune Robotics, which builds autonomous underwater systems for ship cleaning, by Sequoia. Omnea, AI procurement infrastructure, by First Round. The verticals span every category YC has ever published an RFS about, and several it has not.

The implicit argument is that if you find the right people, the territory will reveal itself. You do not need a map.

How these opposing models are shaping private markets

Both models work. The data on YC is unambiguous: roughly 4.5% of YC companies become unicorns versus a 2.5% baseline for venture-backed seed-stage startups generally, and around 45% reach Series A versus a 33% average. EF, using more recent and more European numbers, has produced a $16 billion portfolio at a fraction of YC's deployed capital.

The interesting question for investors is which approach predicts the future better in a market where AI has compressed the cost of building software by an order of magnitude, where defence and dual-use technology have become permanent categories, and where the bottleneck on the next generation of companies is increasingly unclear. If software is genuinely becoming cheap to produce, the constraint shifts from ideas to people who can identify which problems are now worth solving.

That favours the talent-first model on the margin. It also suggests that reading any "this is what we are funding next" document, including a YC RFS, is most useful when it is treated as a description of where momentum already exists, rather than as a forecast of where it will form.

The pattern matters at our level too. The companies that look most like generational outcomes from a 2030 vantage point are rarely the ones that fitted neatly into a published list. They are the ones whose founders were already circling a problem before anyone else thought to write the problem down.

The bottom line

YC's Summer 2026 list is the clearest signal yet that AI is now the substrate, defence is now a permanent line item, and space is now infrastructure. As a description of the present moment, it is sharp. As a forecast of which specific bets will pay, the historical record suggests roughly half the categories will look prescient in 2030 and roughly half will be quietly removed from the next iteration, replaced by whatever the partners failed to see this time round.

Five thousand miles from Mount View in Silicon Valley, Entrepreneurs First continues to operate without publishing a map at all. Its founders move from London and Paris to a residency in San Francisco, find each other, find a problem, and only then find capital. The first artifact is not an idea, but a person.

In a market where the cost of execution has collapsed and the cost of judgement has not, that ordering may turn out to be the more durable one.

What we’ve been working on at Shuttle

• Fundraise is live and momentum is building 💰

• Implementing preparations for our UK launch 🚀

• Shaping our go-to-market plans for the year ahead 👀

The Unsophisticated Investor is brought to you by Scott & Rob, the founders of Shuttle. We’re both sick of private markets being a playground exclusive to the ultra-wealthy so we started a company to challenge the status-quo. Shuttle’s singular focus is to unlock private markets for Millennial and Gen Z tech professionals and help them build wealth through the highest performing private market opportunities.

Scott & Rob
Shuttle Co-Founders