The difficult thing about the AI trade is that both outcomes are brutal.

If AI delivers on its terrifying promise (I.e. real superintelligence, or something close) we lose our jobs. Not some of them. Most of them.

If it doesn’t (i.e. if LLMs are a dead end and scaling laws hit a wall) then trillions in investment evaporate. Markets crash. Portfolios crater.

Either way, there’s tremendous loss and devastation.

The hope is a golden mean somehow, somewhere. AI powerful enough to matter, but constrained enough to augment rather than replace. Productive enough to justify the capital, limited enough to preserve human agency.

I’m not sure that middle exists. But the only way out is through, and the only way through is to keep at it.