If you want an unsustainable joyride, then you too can work for a startup headed for chapter 11 (like 8 out of 10 startups).

Also, knowing people in general has never not been valuable. It's arguably the main thing holding most engineers back from success. Plenty of evidence backs this up: Type fewer keys, press more palms.

Their startups get acquired, or they get acqui-hired because of the founders and core people, who in turn have some pedigree which send strong signals.

If you look at the pattern, it is roughly:

- Top ML/AI researchers from the past 10-15 years - Top ML/AI engineers from top labs or companies

Those above people were hired by FAANG companies or AI startups.

But most important of all, it isn't a overnight success. Many of these have been going at it for 10+ years now. Of course some are relatively young, as in having joined Anthropic or OpenAI more recently (past few years), and launching their own startup after 2022.

This is the way tech has always been. Or as Carlin said "It's a big club... And you ain't in it."

Occasionally, you can ride their coattails (there will always be rising tides), but if you don't have the right connections and the right degrees from the right universities, you're not invited to the box seat TLA titles of these things...

The odds are much better than state lotteries, and the payout is higher than the stock market, but most of us still need to do our own financial planning instead of socializing the risks of the founder class as they privatize the gains. And yet, each day I get these ludicrous emails from a16z about which startups started by which all star founders wants to hire me for less than I make from passive revenue these days for bread crumbs of equity.

But the thing here is, who needs character.ai in an age of local models and cheap tokens from OpenAI and Anthropic (unless you're very naughty and not shaming, just noting)? High margin role play disrupted by cheap open weight models: how very Clayton Christensen.