Capital is an instrument, and the word flattens a dozen different instruments into one. A check from a megafund is not the same as a check from a solo GP. An angel writing from personal net worth is not the same as a rolling fund deploying quarterly commitments from LPs. A syndicate of two hundred small investors on AngelList is not the same as a Series B led by a firm that needs a billion-dollar outcome to move the needle on its fund. Each instrument has its own incentives, its own time horizon, its own definition of success, and its own definition of failure. Before a founder takes money, the founder should be able to explain in one sentence why this instrument fits this company at this stage. If that sentence does not exist, the wrong capital is about to be accepted, and wrong capital is more expensive than no capital.
Already working is not a revenue number. Revenue thresholds are lagging indicators, and founders who chase them before raising end up optimizing for the wrong things. Already working means something narrower and harder to fake. It means there is a repeatable motion that produces an outcome you can predict. It means you know, within a reasonable band, what happens when you spend the next dollar. It means users who were not convinced by you personally are using the product and returning. It means retention cohorts that do not collapse. It means a channel, even a small one, that you understand well enough to buy from. If you can describe the machine and the machine runs without you standing next to it, capital will accelerate it. If you cannot, capital will paper over the gaps and you will raise again in eighteen months having learned nothing except how to raise.
Venture capital is built around a specific shape of return. The math is a power law. A single fund-returning outcome pays for the portfolio, and everything else is noise. The research on median fund returns is not flattering: most VC funds underperform the S&P over long horizons, and the top-quartile dispersion is where the entire industry's reputation lives. That is the structure, and it produces predictable behaviors: FOMO dynamics at the top of the market, pressure to exit rather than build durable cash-generating businesses, misalignment when a founder wants to run a great eighty-million-dollar-a-year company and the fund needs a ten-billion-dollar outcome. None of this is a reason to refuse VC. VCs are good at specific things: network density, follow-on capacity, governance experience, and pattern recognition built from watching many companies fail in many ways. The relationship is not adversarial when both sides are honest about the instrument. It becomes adversarial when the founder pretends to be building a venture-scale company to get the check, and the VC pretends the founder's business is venture-scale to deploy the fund.
Angels are structurally different. An angel writes from personal capital, does not have LPs to answer to, and does not need to return a fund. This means an angel can afford to back a company that will be good but not enormous, and that changes the relationship at the cap table. Micro funds, roughly five to fifty million dollars, operate on similar mechanics. A fifty-million-dollar exit meaningfully returns a ten-million-dollar fund. The same exit is a rounding error on a billion-dollar fund, which is why billion-dollar funds push portfolio companies toward swings that would otherwise be unreasonable. Solo GPs and small teams have grown substantially since 2020, and the evidence suggests this is not a cyclical blip but a structural response: smaller vehicles can back smaller but still excellent outcomes without distorting them.
AngelList rewrote the plumbing of early capital. Syndicates let a lead investor aggregate smaller checks into a single line on the cap table. Rolling funds let GPs raise on a quarterly subscription basis instead of the traditional decade-long close. The cap table of a 2026 seed round looks structurally unlike the cap table of a 2010 seed round. Newer vehicles like USVC point further in this direction: pooling capital from individual investors and deploying it into private companies that those investors could not access alone, letting retail-scale money act with institutional reach. The exact mechanics of any specific vehicle will evolve, and founders should read the actual terms before celebrating the trend, but the direction is unmistakable. More people can invest. More founders can be funded. The intermediary layer is thinner and the capital pool is wider.
Nobody knows exactly what replaces the 2010-2020 model. What we can say is what is already happening. AI is collapsing the capital required to build a working company. Teams of three are shipping what teams of thirty shipped five years ago. When the cost of building drops, the waterfall shifts down. Companies that used to need a seed round to exist can now be started on an angel check. Companies that used to need a Series A can reach meaningful scale on a micro-fund seed. The traditional VC staircase was calibrated to a cost structure that no longer holds. The 2/20 model is under pressure from solo GPs charging less, from rolling funds changing the liquidity assumptions, and from the simple fact that many great companies of this decade will not need the check sizes the model was built to deploy. The direction is smaller checks, earlier bets, more distributed capital, and less dependence on any single intermediary.
A founder in Berlin or Lagos in 2026 has the same tools as a founder in San Francisco. Same models, same infrastructure, same distribution channels, increasingly the same access to capital. Capital is still geographically clustered and Silicon Valley still has advantages in density and follow-on. But the number of investors, founders, and companies is growing across the globe at a rate that any single hub cannot absorb or contain. Capital is catching up to where builders actually are, because it has no choice.
Signing this rule means you commit to treating capital as fuel for something that is already working, never as a trophy, never as validation, and never as a substitute for a business that can stand on its own.