I am listening to Sam Altman on David Senra and it is clear his background as a startup investor influences how he runs open AI

He makes an analogy from allocating economic capital into venture bets to allocating economic/technical/entrepreneurial capital into research bets. The main idea is the power law: a few bold bets that work make up for all your bold bets that don’t. You are in a market of options with positively skewed payoffs.

David offers a view that he is quite similar to Steve: maniacally obsessed about product and figuring out how something should feel based on personal taste. Sam pushes back, which leads me to this idea

There are two main models of founders. The first is product led. They have a taste for what the customer wants and tailor the design of the product/service to be as good as possible. Their edge is largely human and they are designers at some level.

The other is the investor. There is a mass amount of capital they have attracted (money, people, brand) and they need to allocate it to maximize ROI. they have great instincts on the shape of good bets and also re allocating when a bet isn’t paying off. These founders aren’t the designers themselves, they’re closer to operators.
—
The best thinkers do not have general principles

Everything is case by case based on independent first principles thinking

Maybe a product of not being able to predict them. A predictable thinker follows general principles by definition. Unpredictable but not random cannot be boxed into principles or at least principles you cannot model. So perhaps they have general principles but in your view they don’t, in the meta view they do