on founder risk
There's a structural problem with how Silicon Valley has designed the founder experience, and William Hockey puts it plainly: we've made starting companies too safe. If you're moderately competent and went to the right schools, you get a seed round. Worst case, you fail, the company shuts down, and you land a senior role somewhere with 'founder' on your resume. Life is fine. The system has produced a lot of founders who are, in practice, employees who like the title.
The asymmetry he points to is real. An early employee takes more concentrated risk than a founder in most early-stage companies. A 24-year-old leaving Google to be the third hire is giving up $400k in comp, can no longer buy a house, and is betting four years of their life on an outcome they can't control. The founder, meanwhile, will almost certainly raise the next round regardless of company quality, can sell secondary at some point, and if it fails, has a CEO title to trade on. We've de-risked the founder and left the early employee holding the actual exposure. Hockey's argument isn't that we should de-risk the early employee — it's that we need to re-risk the founder.
The version of this he lived: the Visa acquisition of Plaid was blocked by the DOJ in 2021. He had no liquidity. He pledged over a billion dollars of illiquid Plaid stock to borrow $70 million, bought a bank, spent three years near bankruptcy, and got margin called multiple times before the company was profitable enough to service the debt. The fear and the constraint, he says, created a different kind of thinking. When there's only one door, you go through it. The creativity and intensity that produces isn't something you can manufacture in a well-funded safe environment.
What I find genuinely interesting about this argument is the relationship between constraint and conviction. The long-term bets Column made — buying a regulated bank during the Biden administration, not raising external capital, building an employee model built on annual liquidity rather than ten-year illiquidity — those decisions required a founder whose thesis was the only option, not one of several hedged positions. The bet had to work because there was no fallback. That's a different category of commitment than most companies operate under, and it shows in the decisions.
The broader Silicon Valley critique underneath this is sharper than it sounds in polite conversation. A lot of companies that exist are optimized for raising money, not for building the thing. The fundraising cycle warps strategy: stablecoins become a priority when stablecoins are fundable, AI wrappers get built when AI wrappers are fundable. The strategy bends toward what the next round looks like rather than what the company should actually be building. Hockey describes it as the hamster wheel — structural, not a failure of individual will. The mechanism makes rational actors do locally rational things that are globally suboptimal. Getting off it requires not getting on it in the first place, which is only possible if you have a reason not to need it.