eri

she • computers, AI & biohacking • tranny butchdyke left-accelerationist • FALGSC

[Nia Frome @NiaFrome]

imagine having the gall to call xi motherfucking jinping "a broken clock"

[tenso @distributedkv]

the one who gets AGI in the hands of people not controlled by any single entity will win the Nobel prize

[Dr Singularity @Dr_Singularity]

The anti AI phase will end the same way every anti technology phase ends:

with mass adoption

[eri @femaphore]

This just moves the issue one layer of abstraction deeper. Many people, for various reasons, have poor trust in the American government specifically, and this extends to its regulatory bodies, and certain aspects of these policies are unaffected by elections to the best of our knowledge.

[eri @femaphore]

@HannesThurnherr @anonymoustmuki @RoyalCities @mooncat_is This entire argument falls on the assumption that A\ is a good actor. An opaque good actor. An unauditable good actor. And that it will stay that way for as long as they hold the frontier. That is an insane assumption to make.

[eri @femaphore]

@eepyeri @CappyIshihara ancient tricknology

[eri @femaphore]

so you're telling me Claude Fable 5 disproved the yakubian conjecture. interesting.

[eri @femaphore]

@lilypeonyyy @CappyIshihara And yet, humanity always finds a way to make technology more sustainable and efficient. Year after year. But no, this one is special. This one is ontologically evil. This specific kind of matrix multiplication will always be this way.

[eri @femaphore]

@lilypeonyyy @CappyIshihara Yes, so am I. How long has it been, maybe three days since Fable 5's latest breakthrough?

[eri @femaphore]

@lilypeonyyy @CappyIshihara AI does not boost productivity and does not give you new information, it doesn't do shit.

Guess Torvalds simply has AI psychosis. Guess the field of mathematics isn't having daily breakthroughs, it's just illusory. Guess pharmacology isn't discovering new compounds. Oh well.

[eri @femaphore]

@lilypeonyyy @CappyIshihara so we're just making shit up now, why is it always like this with you people?

[eri @femaphore]

@lilypeonyyy @CappyIshihara neo-luddism is not the same as luddism.
Luddites were against labor displacement and exploitation. This is a reasonable position.
Neo-luddites are simply against technological progress, disconnected from its labor impact.

Hence why I asked, would it end, even under net-zero?

[eri @femaphore]

@lilypeonyyy @CappyIshihara caps don't make an argument more meaningful

[eri @femaphore]

@lilypeonyyy @CappyIshihara we can probably revisit this conversation when inference is as efficient as public transport is, then.

[eri @femaphore]

@lilypeonyyy @CappyIshihara and let's be real here, would this neoluddite opposition end, when inference is as efficient as the innumerable other technologies that have had decades upon decades of iterative improvement?

[eri @femaphore]

@lilypeonyyy @CappyIshihara no one's taking your hobbies away. you can still do calligraphy even though the printer has existed for decades.

environmentally? fair, but not really an argument you're morally entitled to make unless you're a vegan, don't drive a car, etc.

[eri @femaphore]

@lilypeonyyy @CappyIshihara so again, what is the opposition to open-weight models?

[eri @femaphore]

@lilypeonyyy @CappyIshihara "taking the humanity from humans" is a meaningless platitude.

[eri @femaphore]

@lilypeonyyy @CappyIshihara the premise of your position is incorrect.

AI can be, and already is, a replacement for human labor. under capitalism, this displaces labor, and is worth opposing; but the root problem continues to be capital, not automation.

[eri @femaphore]

@lilypeonyyy @CappyIshihara so the latter

[eri @femaphore]

@lilypeonyyy @CappyIshihara so, you'd be fine with an open-weight model with ablated guardrails, right? right????

or maybe it's just vibes-based ethics