no notes

[saxnot@chaos.social]

@orange_lux @dibi58 @karl yeah so?
we had recommendations long before our current gen (pre llm) stuff

so?

data science is a trillion dollar industry
yeah there was more than one generation of recommendation algorithm
llm are a new tool and at some things they are better

[saxnot@chaos.social]

@orange_lux @dibi58 @karl those amazon recommendation matrices are not cheap either

not really a comparison that is fair but LLM is just a minor evolution in a field with centuries of real life usage, development etc

[saxnot@chaos.social]

@orange_lux @dibi58 @karl today I don't have the capacity for whataboutism

you're missing the engineering point

there is already enough misinformation out there

[orange_lux@eldritch.cafe]

@saxnot @dibi58 @karl so let's end this conversation here, then.

[hex@kolektiva.social]

@karl @sherbang @orange_lux @saxnot @dibi58 I know how much it takes to run at least basic models because I'm running local models for experiments. I won't use hosted models because I'm not giving them money or training data. But yeah, capitalists are trying to sell something that doesn't exist and that they don't understand.

I would be entirely unsurprised to find out that even inference on corporate models can't cover costs. To sell "AI" it has to be a thing that just works for everything all the time. It must take no thought. That's incredibly wasteful.

It needs to stop being subsidized, just like cars.

[tuutw@mastodon.nl]

@karl En Politico zal ons even vertellen wat wij moeten doen of laten? And Politico is the instance telling us what to do or not?

[yenndc@kolektiva.social]

@karl

Reference, for whoever might care…

https://www.politico.eu/article/europe-choose-ai-climate-goals-data-center-chief-warns/

[a_goodall_spaceship@norden.social]

@cucufaiter @karl 9 minutes? Yep same with me, spent all my tokens, query still running.

[cucufaiter@mastodon.social]

@a_goodall_spaceship @karl miliseconds xD

[saxnot@chaos.social]

@Hex @karl @sherbang @orange_lux @dibi58 100 % agreed

the AI hype is absolute madness
can't wait for the bubble to burst

[jens@social.finkhaeuser.de]

@karl Clear choice there, except for absolute idiots who think they can exist without the environment their species is adapted to.

[nini@oldbytes.space]

@dibi58 @karl Then more focus on machine learning, less on large language models? I agree, LLMs are a hideous waste by design.

[nini@oldbytes.space]

@rumpelheinz @karl Ein gutes Argument dafür, warum wir keine KI brauchen.

[karl@infosec.exchange]

@nini @dibi58 there's an in-between where you still train an LLM but it's more focused, but in essence, yes.

[nini@oldbytes.space]

@karl @dibi58 Can't see why LLMs need to be involved, they're not very good and the sheer amount of effort in training an LLM from data collection to tagging to getting anything useful out of it is in opposition to any climate goals, moderate as even the most radical ones proposed will be.

[karl@infosec.exchange]

@nini @dibi58 I would like to see research supporting such an assertion.

I dislike LLMs very much, which I why I don't want to make unsubstantiated claims that can be debunked by LLM supporters, which would discredit any other concern, however valid, that I have with LLMs.

[dc4dd@sueden.social]

@karl Politico is Springer Press

[lispi314@udongein.xyz]

@orange_lux@eldritch.cafe @dibi58@this.mouse.rocks @karl@infosec.exchange Technically, given the kind of compute for energy ratios that human brains (and other dynamic nanosystems) manage to accomplish, it should be quite possible to accomplish with synthetic biology.

We just need to get much better at it and forget ever accomplishing it on static compute (stop even trying, the efficiency is too bad).

The efficiency of digital static compute is laughably bad at the best of times.

[uriel@bbs.keinpfusch.net]

@lispi314 @orange_lux @karl @dibi58

> synthetic biology.

we don't really need this. There are so many bacteria and neural cells available, with billions of  years of evolution...

--
Uriel Fanelli
Using Aktor: https://git.keinpfusch.net/loweel/Aktor-2
XMPP: uriel@keinpfusch.net
old blog: https://blog.keinpfusch.net
new blog: https://keinpfusch.net

[lispi314@udongein.xyz]

@ALewpqX9B5hLrwnPl2.uriel@bbs.keinpfusch.net @orange_lux @karl @dibi58 That's a question of degrees at that point.

Even if you use building blocks that already exist, isn't it synthetic?

Similar to a hammer, all the components remain natural (since they clearly exist to start with), but intelligent intervention was involved in its making. That at least is the distinction I draw.

There's no particular reason to recreate it entirely from scratch unless you very specifically want to make it deliberately incompatible with the rest of the biosphere.