David Gerard

When anyone says "LLMs as normal technology" see if it's actually any better than "blockchains as normal technology" or "NFTs as normal technology"

this is the excuse crew trying to calm the increasingly pissed off crowd.

"AI" is not a "normal technology". it is a *political project*. its proponents are extremely explicit about their goals with it. the people claiming we should just treat this political project as "normal technology" are shilling.

Oct 22 at 12:37amWeb
Show threadColinOct 22
@davidgerard Being a useful idiot is very normal for humans.
Show threadwakingrufusOct 22
@davidgerard
I'd be open to this framing if anyone who used it actually followed through. Sure, so we burn enough energy to power a town and/or pay more money than all of our commercial IDE and build tool licenses combined, and what we get is auto complete that can guess in bigger chunks, but is sometimes completely wrong. By any "normal technology" metric this would never be worth it.
Show threadSu_GOct 22

@wakingrufus
Great description of (so-called AI) LLMs!: “auto complete that can guess in bigger chunks, but is sometimes completely wrong.” 😁
And so true: “By any "normal technology" metric this would never be worth it.”😐

@davidgerard

Show threadTofu ​Oct 22
@davidgerard
Show threadJacquelineOct 22

@davidgerard

LLMs is, at best a flashy tech demo, and at worse a deliberate project of neofascist billionaire elites to poison the information pool and destroy the promise of public knowledge and free association on the Internet.

Show threadStarkRGOct 22

@burnoutqueen @davidgerard LLMs are unethically-produced, environmentally-destructive toys with no legitimate use that isn't already handled by something better, more ethical, and more environmentally friendly.

They're the equivalent of easy bake ovens that run exclusively on coal, are manufactured by slaves out of toxic materials, and only half-bake everything.

Show threadJacquelineOct 22
@StarkRG @davidgerard but but what would the slop manufacturers use to meet their quotas for bogus information?
Show threadStarkRGOct 22
@burnoutqueen @davidgerard They managed pretty well for centuries before LLMs existed, they just had to pay ethically deficient people to do it.
Show threadAsta [AMP]Oct 22
@davidgerard “sure, the wastewater is tainted with mercury… except NOW I’m out here using that as a rich supply to make and sell mercury thermometers!!”

“but we already have better, safer on-“

“WHOA THERE Captain Luddite, how dare you, my mercury thermometer business is just about to enter the first stage of investor funding. WOW. Idiot. Jerk.”
Show threadtdietterichOct 22
@davidgerard The normal tech argument is intended to counter the hype that LLMs are going to become super intelligent. If you read the original essay, the main claim is that, like other new technologies, it will probably take 20 years to figure out how to use it appropriately.
Show threadJenica Lake เหล็ถเลถOct 22
@tdietterich @davidgerard LLM's are psychological warfare, propaganda machines, nothing more, but nothing less.
Show threadtdietterichOct 22
@davidgerard AI (as a long term effort) was originally motivated by philosophical, scientific, and mythological (eg Promethean) concerns. But it rapidly became a marketing term as well. In my experience, it is only in the last 5-10 years that a significant fraction of AI researchers began to appreciate the political aspects. Of course they were always there, but now they cannot be ignored.
Show threadDavid GerardOct 22
@tdietterich and yet I see it used to say "ignore the politics".
Show thread@iveylineOct 22
@davidgerard So true.
Show threadAmgineOct 22

@davidgerard

Nonsense. This is a pump-and-dump scam, like (almost) everything corporate related to computers.

The winter nudists in the flock will shiver, some will die, same as usual.

Show threadBenjamin ReedOct 22
@davidgerard @daedalus Yup. “LLMs as normal technology" have already existed for decades, it’s called ML and it doesn’t involve putting the entire text of the internet into one giant incredibly inefficient model.
Show threadFoolishOwlOct 23

@davidgerard I continue to recommend Resisting AI: An Anti-fascist Approach to Artificial Intelligence, for reasons the title makes obvious.

https://en.wikipedia.org/wiki/Resisting_AI

Resisting AI - Wikipedia

Show threadJonOct 24

yeah, great book.

@foolishowl @davidgerard

Show threadSindastra :femin:🏴‍☠️Oct 24
@davidgerard I don't follow AI stuff much so I'm a bit out of the loop. Can you explain a little more? ​​
Show threadravendarkholmeOct 24

@davidgerard As a species, we have become lazy and self inflicted ignorant.

The last decades we have been collectively taught that we need to persue our happiness no matter the cost, even if it's at our own peril.

Masses want entertainment and new wonders like Hollywood promised us,even if that means creating something that will eventually destroy us as we knew ourselves before AI.

People will look backwards to the world we are about to loose because of our bottomless egomaniac desires and compulsions which have destroyed this once beautiful planet 🌎 🌏 🌐

Show threadJonOct 24

https://neuromatch.social/@jdp23/115418779049342835

@davidgerard

Jon (@[email protected])

And your claim we all want AI to be viewed as a "normal technology" in the sense Narayanan and Kapoor are using the term is also a claim that we all buy into the centrist position. From [A guide to understanding AI as normal technology](https://www.normaltech.ai/p/a-guide-to-understanding-ai-as-normal) >" In short, if you are looking for a moderate position, we encourage you to read the essay in full. Don’t let the title fool you into thinking we are AI skeptics. Perhaps you will conclude that AI as Normal Technology is already the middle ground you are looking for." What about those of us who *aren't* looking for a middle ground? @[email protected]

neurospace.live