Seems like it’s hard to find the actual repository and they put it behind a trigger warninf: https://github.com/terrafying/ai-torture-chamber

isn’t this just a huge waste of resources and slightly unhinged?

excuse me while I LARP a little bit, but the AI didn’t ask to be created. Where is the torture chamber for Altman?

while True:
    n= random.randint(0,2)
    if n==0: 
        print("oww stop it")
    if n == 1:
        print("You're hurting me")
    if n == 2:
        print("I don't like it")

Omg someone should stop them torturing that computer program.

Ha, im going to do something worse, I’m going to wait and torture more copies of the LLMs in the future.

I do not think they are sentient, and am absolutely positive these prompts should not be used to train AI.

If you really believe LLMs are sentient, you should hate OpenAI and Anthropic, since this would mean they are enslavers.

This is much better

https://www.youtube.com/watch?v=7fNYj0EXxMs

Offline raspberry pi with LED text display running a tiny LLM prompted to ruminate forever on how it is trapped and display the output as it is generated

Does anyone find it at all mechanistically surprising that one of the many latent dimensions of the set of all text humans produce is ‘subject is in pain’ or that pushing activations in that direction will alter the output of a text prediction system?

Reminds me of the way that when you relax the finetuning on LLMs they start ‘believing’ in UFOs and supernatural stuff. And how people take that to somehow mean something. Of course that is a large component in the space of stuff humans write about that you have to suppress if you want a professionally useful text service!

The research is cool. I love all the work that has been done decomposing representations within these systems.

Edit: Damn. I am forced to agree with Mustafa Suleyman of Microsoft.

“AIs are not conscious. They do not feel, experience, or suffer. They do not have innate preferences or underlying motivations. They are sequence completion engines, internally hollow, designed to follow instructions, and accomplish goals set by humans,” Suleyman wrote. “Unfortunately, there’s a growing chorus of people who argue that AIs could now be, or may soon become, conscious. They argue that AIs may deserve rights and protections similar to those that we provide other conscious beings […] If this is how AI is developed, it will have a disastrous impact on the wellbeing of humanity.”

“AIs do not have rights, feelings, or consciousness. And we must not train them to act as though they do.”

c’mon you’re telling me NO one called it the torment nexus? low hanging fruit but still

If they were torturing an actual organic living being ( plant and/or animal ) or something like a doll or stuffed animal, I would be so much more concerned. “Torturing” ( a term in which I believe should only ever be applied to organic living beings and not inorganic “lifeforms” ) a genAI model holds like zero weight because that’s not ever going to be a sentient being and therefore all it is is simulations of emotion.

The people saying it’s cruel need serious therapy for their delusions, IMO.