The piece (and lots of commentary) keeps talking about "AI Companies" as though it's this fixed set of companies that are destined to all be the same. But anyone can start an AI company... the models are all available, and the surrounding technology is fairly accessible. Yes, there's a lot of companies that will always maximize engagement. But... anyone could make something that doesn't serve that goal.

▲

Analemma_ 4 days ago | parent [-]

That's not a stable equilibrium though: if you don't maximize engagement, you'll be outcompeted, outspent, and probably ultimately acqui-hired and "our incredible journey"-d by the companies which do.

And if you don't think this is the inevitable outcome, note how every social media platform has, slowly or quickly, gravitated toward maximizing engagement to the exclusion of all other priorities. Why would AI be any different?

	▲	ianbicking 3 days ago \| parent \| next [-]
		"Why would AI be any different?" It might not go any differently, but I think all us folks here have an opportunity to make it different. Any product needs to pursue enough engagement that the user actually gets value. I have a bunch of apps I installed aspirationally, but don't use. But if we're willing stop ourselves once we have enough, we can pursue enough engagement and enough revenue. But if you tie yourself to someone or something who will never have enough then I agree, you'll get sucked in.
	▲	senko 4 days ago \| parent \| prev [-]
		Because social media is a winner-take-all with strong network effects. AI isn't.