There’s an awful lot of people who love to think about the grey-goo problem of nanotechnology because it’s flashy and sci-fi who don’t want to think about how LLMs are capable of doing that to the Internet. - ThreadSky

cooperlund.online • 2 days ago

There’s an awful lot of people who love to think about the grey-goo problem of nanotechnology because it’s flashy and sci-fi who don’t want to think about how LLMs are capable of doing that to the Internet.

Reposted from GOLIKEHELLMACHINE

i do think there's sort of an interesting business problem emerging for someone like google, because i'm not sure how big the market for auto-generated video content with auto-generated ads could actually be, but i suspect the volume of that content will make it even more unusuable

Comments

golikehellmachine.com•2 days ago

the thing you see over and over again when you watch execs at the big companies talk about this stuff is how little interest they have in what business problems all this shit solves, and they have even *less* interest in what business problems all this shit may cause *for them*

bryyyyguy.bsky.social•2 days ago

I’m often wondering what the tech lords think will happen if AI can in fact automate away millions and millions of jobs. Those “email jobs” spend money all over the economy and I don’t think destroying what’s left of the middle class standard of living will do good things to social cohesion.

nachyocheese.bsky.social•2 days ago

That's why they all have bunkers

golikehellmachine.com•2 days ago

most of the ones who have thought about it assume that all of these people will become their feudal peasants, a few of them think we'll get butlerian jihad, most of them don't think about it at all

lizamazel.bsky.social•2 days ago

I hope they like big fortified castles behind 15 foot razor wire wrapped gates and rarely leaving them

which "Living Dead" movie was this one, again? after the mall?

tampalib.bsky.social•2 days ago

He just like me fr

But seriously, it's the exact same behavior you see among us common folk spitting between AI boosters and AI doomers. We can't abide the idea that it's a real technology with real mixed effects that are already starting to manifest.

pemo.bsky.social•2 days ago

And LLMs increasingly fed with LLM generated content might become (even more) useless much faster than one might anticipate:

https://arxiv.org/pdf/2305.17493

pemo.bsky.social•2 days ago

„What will happen to GPT-{n} once LLMs contribute much of the language found online? We find that use of model-generated content in training causes irreversible defects in the resulting models, where tails of the original content distribution disappear.
We refer to this effect as model collapse'“

Comments

Posting Rules

Reply