Thread | Brutkey

"The way I see it, during any given period the market is driven by a narrative, positive or negative. The current narrative is strongly driven by optimism about AI. And it takes a major shock to change a market narrative.'

And Krugman goes on to note that investment in AI data centres accounts for about half of all US economic growth in 2025.

AI winter is coming. And then there will be a depression. (NB: that's me, not Krugman.)
https://mastodon.social/@rbanffy/114981848940676725

redsakana
@redsakana@infosec.exchange

@cstross@wandering.shop It appears that a stopgap bullshit hype object has already been prepared: "Quantum". IIRC even Ursula von der Leyen has been bringing it up lately.

Compared to the previous hype cycle, it seems to me that "AI" roughly equals Bitcoin: problematic in many respects, but has technically interesting aspects, and more or less works for some applications (though not the most-hyped ones).

"Quantum" would then be the new NFT: to anyone with half a brain obviously 100% scam from the start. But there are a lot of retail investors with less than half and even more people who would like to rip them off.

Poligofsky 🇨🇦

@8r3n7@mstdn.ca

@cstross@wandering.shop

Is there any hope that the computation capacity can be retooled? How different are AI chips compared to other vector parallel processor GPUs?

Given a massive quanitity of compute power, what are some better ways it might be deployed? Virtual reality worlds is an obvious option. AIs could serve as better NPCs, and do other menial tasks that players and staff can't or won't.

But surely there are other options? Running scientific simulations? Virtual biological and genetic experiments?

Is it wrong to assume that a more powerful civilization will require more computation? But that it's a question of the best applications?

David Penfold :verified:
@davep@infosec.exchange

@cstross@wandering.shop another AI winter. Unlike the last one, this is going to have devastating economic repercussions.