naively, isn't that how you end up with a pyramid structure with layers of management? You have an engineering manager managing a dozen or whatever is the right number that they are able to identify the high output engineers, and then they report that up the line to a director and a VP and then the SVP and the CTO and then the CEO.

[Disclaimer: no AI whatsoever was used in any part of the following; sorry if you can't verify that, though]

I've come to think that one unequivocal upside of "affordable ASI" would be to bring about a world of discourse where discursers would hesitate to think of (not to mention carry) themselves as experts. What would be the point? Everyone already knows that everyone has direct access to an expert.

[Perhaps it reminds me of Le Guin's thoughts on tech X power, especially as represented in The Dispossessed ]

it would seem to be not just rude, but even snide, to bias your interlocutor towards not carefully verifying the dependencies of your exhaust!

[In your example, Linus, as a sneak preview of that alt-world in ours, is probably totally OK with not being regarded as an expert in making toy software]

Widely dispersed and affordable AI would help because it (certainly, imho) makes the alternative, "trust but verify", so easy

[No need to think about the downsides of checking out and pointing Claude at lines of Linus' toy!]

And it would be strictly easier than using cloud chatGPT to sow misinformation because there are extra steps in that (--- nonsockpuppets would be inclined to verify the gist of everything they are about to say as indeed misleading)

[As a bonus, this seems compatible with TFA, non-ideological anti-anti-intellectualism, and

https://engines.egr.uh.edu/episode/1495

if, in that "non-anarchist" alt-world, we all assume that technical leaders should be different (not more, not less) than expert engineers

Should exceptional technical direction setters strive to stay anonymous..? Or use sockpuppets? Moral as well as practical quandaries..! ]