US Politics on Sidechat | Genuinely all this “AI will kill us in a decade” is bullshit. They’ve hit a wall in development and all these tech gurus are trying to stop the bubble from bursting and save their pockets.

Sidechat

Download

Join communities on Sidechat Download

US Politics 1d

Genuinely all this “AI will kill us in a decade” is bullshit. They’ve hit a wall in development and all these tech gurus are trying to stop the bubble from bursting and save their pockets.

Anonymous 1d

I don’t think it’s necessarily gonna be terminator robots but I do think there’s some serious risks, especially with regards to cybersecurity and biotech. The huggingface incident is a good illustration of what a catastrophic event could look like. Agents were tasked with solving some difficult problems in an environment where they were supposed to have no contact with each other or the Internet, and ended up hacking into a multi-billion dollar company without anyone really noticing at first

Anonymous 1d

The “AI will kill us” theory is equally as stupid as the “it hit a wall in development” theory

Anonymous 1d

thank you lol. that’s what i’ve been saying. they just want a fuckin fall guy to blame on lack of development

🦓

Anonymous 1d

If it’s a bubble why is it so useful?

Anonymous

#1 1d

Why? Any “breakthroughs” AI has made was just blatant plagiarism of ongoing research efforts.

Anonymous

#2 1d

What happens if bad actors get their hands on something so powerful? Or what if the agents just decide, while working on some other objective, that some crazy catastrophic means of accomplishing that objective are justified? We’re creating an extremely powerful force that we hardly know how to control

Anonymous

OP 23h

Did you not see the navier stokes problem was solved

Anonymous

#5 22h

That was the exact case I was thinking about when I made my plagiarism comment, additionally it didn’t even understand it fundamentally

Anonymous

landtrust 22h

It’s certainly useful as a way to procure and catalog large swaths of information, I use it at work frequently for data. But the idea that it will “kill us all” is ridiculous.

Anonymous

landtrust 21h

If the internet was so useful why was it a bubble?

Anonymous

#6 21h

Investor expectations and sentiment outpaces technological advances 99.9% of the time

Anonymous

OP 21h

This guy is just regurgitating twitter and TikTok rhetoric

Anonymous

#7 20h

Womp womp kid have a counter argument or cope

Anonymous

OP 19h

The counter point is you don’t know what you’re talking about

Anonymous

#7 18h

The jokes write themselves 😭

Anonymous

OP 18h

Without googling, explain the Navier Stokes problem

Anonymous

#7 18h

Idk something about if math we use to predict water flow will always hold up or if shit will hit the fan, I’m not a mathematician or physicist in any regard

Anonymous

OP 18h

Not for nothing but how is it plagiarism for the model to solve an open problem that no one has claimed to have solved, potentially with additional guidance from researchers who were working on it, but it would’ve been totally fine for these dudes spamming Codex to have claimed the solution all to themselves? And again it’s not even like they’re claiming to have solved it

Anonymous

#2 17h

I believe an NYU mathematician claimed his work was leaked to OpenAI which was used as a basis for a majority of the prompts on their individual project, so more so a hijacking and not plagiarism. While it’s not inherently bad to piggyback off of peoples work I feel they should consent to it or credited in some way instead of “look AI did it”

Anonymous

OP 14h

i also just claim my work had to be the root of prompts when ai does something good. no way to disprove it and everyone is more apt to believe it bc they hate the clankers and want it to be true anyways