Winter 2027. Many of the best ideas we've funded surprised us, and several surprised the founders' users. But there are hazards we've been waiting for. If you're working on one of these, apply for W27.
Desmond Achterberg
Every large company will have a whistleblower eventually. Today they find out from a newspaper. We'd like to fund the marketplace where they can find out earlier, from the whistleblower, and make a competitive offer. Think of it as an acquihire of one.
Perry Gault
Founders make commitments (to users, to regulators, to the public) that are fully backed at issuance and then quietly fractionalized. We think there's a large opportunity in tooling that manages the peg between what a company said and what it is doing, so that the break happens gradually and on a holiday weekend.
Marguerite Oyelaran-Pike
Consent is the biggest bottleneck in our portfolio. Users have to read things, understand them, and agree, and the first two steps have terrible conversion. We want to fund agents that read, understand, and consent on the user's behalf. The user is still informed. Their agent was informed, and it's their agent.
Desmond Achterberg
Building a moat used to require years of lobbying and a relationship with a state legislator. We believe a developer-first platform could get a small, badly written licensing requirement passed in all 50 states with a single API call. Incumbents do this by hand. They are ripe for disruption, by you, after which you will be the incumbent and should call us again.
Perry Gault
Litigation finance today funds plaintiffs. But plaintiffs are reactive. The defendant knows years ahead of time that the suit is coming, because the defendant is the one doing the thing. We'd like to see a product that lets founders start financing their defense on the day they decide to need one.
Marguerite Oyelaran-Pike
Indenture (W24) found it surprisingly easy to get a tranche of toddlers rated AA. We think it should be easier. The issuer-pays model is one of the most durable moral hazards ever engineered and it has seen almost no software innovation since 2007, when it had what we'd call a strong proof of concept.
Desmond Achterberg
A great deal of capital is going into making people live longer. People are the wrong target. A person who lives to 150 is still one person with one name. A company can die on Tuesday and be reborn on Wednesday with the same team, the same code, a new name and no debts. We want to fund the tooling that makes this a one-click operation. Internally we call it Chapter 11 as a feature flag.
Perry Gault
Today's children are tomorrow's daily active users, but the market for their attention only opens when they get a phone, by which point it's crowded. Someone should build the exchange where platforms can take a position on a cohort years in advance. We don't know exactly what this looks like. We know Fett (S22) would list 9,000 units on day one.
As always, you don't need to be working on one of these to apply. The best hazards are the ones nobody saw coming, by definition.
