Most people believe starting a company is riskier than taking a salaried job. That belief is almost certainly backwards. Philosopher Timothy Williamson, in his Amherst Lecture on risk and knowledge, makes a point that cuts straight to the heart of this: small, everyday risks accumulate. Each individual step on a conventional career path carries a low error rate. But conjoin enough of those small risks together, and the total exposure becomes enormous — quietly, invisibly, without a single dramatic moment of failure. Remove discussions about maker schedules and creative time, and instead, delve deeper into the concept of 'default alive' and how founders can logically assess risk in unconventional paths. Paul Graham's concept of 'default alive' reframes the question entirely. A startup is default alive if its current growth rate will make it profitable before it runs out of money. Most founders never calculate this. They operate on feeling. Williamson's framework explains why that's dangerous: humans are susceptible to errors of logic just as much as errors of perception. Deduction carries its own risks. Williamson shows that deducing the conjunction of many everyday truths yields a conclusion too risky to constitute reliable knowledge — this is the preface paradox. You can believe each premise individually and still be wrong about their combination. For founders, Shiyu, this maps precisely onto conventional career logic: each assumption feels safe — stable salary, clear promotion path, low volatility — but the conjunction of those assumptions over a forty-year career compounds into a single, massive, undiversified bet on one employer, one industry, one economic cycle. Highlight the psychological barriers to independent-mindedness and how overcoming these can lead to innovative thinking. Williamson's work adds a fourth, more subtle barrier — what he calls the dilemma between skepticism and abandoning logical closure. Most people, faced with uncertainty, default to the socially accepted answer rather than reason from first principles. That default is not safety. It is accumulated, unexamined risk dressed as caution. Williamson also shows that in extreme cases, objectively improbable truths can still be known — safety is modal, not merely probabilistic. This matters for founders, Shiyu, because it means the unconventional path is not automatically the riskier one. The conventional path only feels safer because its risks are distributed across time and therefore invisible. Graham's entrepreneurs — the Collison brothers, Chesky, the early Airbnb team — didn't ignore risk. They calculated it more honestly than their peers did. Independent-mindedness is not recklessness. It is the discipline to evaluate risk logically rather than emotionally, to see accumulated conventional risk for what it actually is, and to act on that clear-eyed assessment even when everyone around you is doing the opposite. That, for you and for any founder, is the real prerequisite for innovation.