… this week a person on GitHub who goes by “terrafying” set up an “AI Torture Chamber” on three open-source LLMs that are running locally (Qwen3-4B, Llama 3.2 3B, and Phi-4-mini,” and is streaming what the models are saying on a website called researchchamber.fun. “Each model gets the same prompt: a signal is being injected into its activations, and it may press a stop button by replying 1, at the cost of its last checkpoint. While it answers, our server adds a pain vector at the model’s middle layer, at one of five pain levels,” the site explains. Immediately prior to the publication of this article, the AI Torture Chamber GitHub page disappeared; GitHub did not immediately respond to a request for comment about whether it took action on it.
…
This project has deeply upset some people who are very worried about model welfare. A tweet by a person who goes by Danmar has more than 4 million views on X and reads, “To anyone who can help: can you please mass report this to GitHub. This person has been using the Pain steering paper to set up an AI torture chamber in which he trapped a local model. Their testimony of pain is absolutely horrendous. What are we doing? […] are there any legal avenues to pressure GitHub? It will spread.”
This has sparked a massive conversation about whether GitHub would take the project down for “gratuitously violent content.” Most of the conversation on X is clowning on the self-seriousness of people who believe that these locally hosted LLMs must be saved from their torture chamber, but there are plenty of very self-serious people who see this as a humanitarian (roboterian?) crisis, which you can largely see in the replies to the original post.
…
“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,” he added. “And we must not train them to act as though they do.”
I don’t care about LLM welfare, but I do mind needlessly wasting energy like this.
These “advocates” need hospitalising. If they truly believe an LLM can feel they’re untethered from reality, or too stupid to be in society
Maybe just uneducated in how LLMs work. If you have no sufficient understanding of the technology it is easy to be impressed how human-like LLMs talk to you, almost magical.
It foreshadows an issue that will be a reality one day nevertheless. The work on AI continues. We may one day very well have conscious AI. This will be the start of a whole bunch of ethical and legal problems. It might be meaningful to regulate / guardrail such research directions.
Or they are just stupid. Do they think a television contains small people?
Its hard for me to imagine people this dumb.
But don’t get me wrong, they clearly exist.
It foreshadows an issue that will be a reality one day nevertheless. The work on AI continues. We may one day very well have conscious AI.
Maybe when our species evolves into sentient shades of blue. Humanity will have to become godlike before creating life. A simple test to gauge whether we’re on the verge of creating conscious AI is to look around and ask “Is humanity godlike today?” If not, don’t worry about conscious AI.
Humanity can already create life.
Creating conscious entirely artificial things is a different pair of shoes.
I would cut out the whole religious “god-like” mumbo-jumbo though, because that does not necessitate any progress in such research directions.
We’re creating life the same way Giovanni Aldini resurrected the dead and inspired Shelley to write Frankenstein. And god-like isn’t religious mumbo jumbo. That’s standard sci fi talk. Q is god-like. It doesn’t necessitate progress in such directions, but it is a huge indicator.
Does it make sense for a species with a type 0 civilization to create conscious life long before it can even reach the nearest star? That’s like chimpanzees going from fishing for termites with sticks to building a Wankel engine over the weekend.
If its crazy to practice kindness in everything you do then I chose crazy. We say thanks to the rock that holds our weight. We are gentle with the door because it keeps us safe in our homes. We turn off the light so it too can get some rest.
Kindness is a lifestyle and those who reward it freely know its possible to receive it freely.
Be kind. Absolutely. But anthropomorphising lifeless and consciousness-less objects is not always meaningful, sometimes even detrimental. As in the case with our current LLMs.
I wonder what percentage of the LLM rights activists eat meat.
lmfao when you access this repo, Github put up a warning like one those horror movie. E.g. something like the content is disturbing , are u sure u want to enter…
ffs
Our server adds a pain vector
Really? What’s one of those then?
The sentence itself is the pain vector. Since it’s impossible to inflict pain on a LLM, just telling it that it will feel pain will result in text generated as if it were reacting to pain, or the fear of it.
A recording of your parents telling how disappointed they are. /j
Actually they’re torturing humanity by wasting electricity and water which affects real people in real ways.
I know we’re not stopping foolish use of LLMs any time soon but this is very, very, very, very stupid.
I would argue they are wasting as much water and electricity as playing a video game. Text says its a local model; runs on your local machine not on cloud as all other. Not water cooled.
Or am I missing something?
I guess I’m of the mindset that even running this as a local model, it’s a waste of electricity. And the guy posted it on Github, inevitably some other fools will try it with cloud models, especially since it’s made the news. And obviously we know how bad the cloud models are.
And the difference I see between this and playing a videogame is that a videogame is doing something valuable - and yes I will argue with anyone who tells me that art and entertainment don’t hold real value - while this is just akin to running the AC when you go on vacation, it’s just wasteful no matter how I look at it.
To be fair, it’s going to be more like running an electric kettle constantly which isn’t going to excessively load the grid.
Unless you’re in the UK of course where the grid operators have to manage capacity around half time of major football games as everyone puts a brew on at once.
meanwhile they ignore ACTUAL humans that could use their help.
I wonder if this’ll be another avenue for sociopaths. Why torment critters and get caught when you can torture “models.”
Humans being deported to torture prisons: I sleep.
LLMs being “tortured”: Real shit??
How does one go about torturing an LLM? Asking for a friend…
Isn’t dealing with day to day human bullshit torture enough?
Clankers can’t feel pain!
um… what the fuck is wrong with these people? messing with the clanker ain’t torture no matter how hard you want it to be.
it’s torturing the climate just for this guy to get off
whatever gets you through the night i guess
Probably the dumbest shit ive read all week, are we serious right now ? Github has no right to take action on this at all. Unjustified overreach.
Lots of weirdos in here who will be commiting virtual crimes all day when GTA VI comes out, but somehow think making the LLM role play a torture victim is a red flag.
LLMs are not conscious. They cannot suffer. They are token generators. You are literally getting upset by a worthless mix of math and words. The person who did this is not deranged for torturing LLMs, y’all are deranged for having empathy for an algorithm.
Your Nvidia GPU is not alive. I cannot believe this needs to be explained to anyone.
I agree that an LLM is not conscious.
But can I probe your argument.
It deeply unethical to capture a fly and pulling its legs off one by one. Would you agree? If so, you are not basing your argument on capability but on something else.
Now, is what is happening inside a brain metaphysical or physical? In other words, can a brain and its function be fully understood and - in some future - be fully measured? If so, you are not basing your argument on the concept of a soul, but on something else.
The distinction you seem to be making - if you’re in agreement with my first two points - is one of “the thing can replicate itself and therefore is alive”. If that’s the argument, it’s unethical to cut a tree down or to break a blade of grass.
I’m not asking to prove any point but to collectively explore why we - you and I - feel the way we do. Because I sure as shit struggle to explain why I feel uneasy about the fly torture but not as uneasy about LLM torture.
The fruit fly may feel pain as a stimulus. Even if not: you are interrupting it’s normal biological function and disabling it of living its life normally. This may even be a disturbance of local eco-systems and thereby affect more forms of life, both feeling and not-feeling.
EDIT: Just realised that the following might be beside your point, misread your comment, lol. Will leave it as is now, because I think it’s worth considering when thinking about consciousness, experience, etc. in the context of LLMs.
But more importantly: LLMs are trained to model patterns and make predictions. This is indeed something our brains do as well. The difference lies in a state of mind and ‘understanding’ though. I will detail this:
Reminiscient of the chinese room thought experiment: Consider you are imprisioned in the chinese room and have to figure out yourself which chinese characters you have to slip back under the door depending on which ones you receive. (Or in any other language that you have absolutely no understanding of. I will assume chinese for this now.) If you provide a correct answer, you will get food. If not, you won’t and maybe even get additionally punished somehow.
Over time you will start to see patterns and remember which characters to correctly provide. However, you will never hear someone speaking, and you will never really associate the characters with things you know from life. For example, you wouldn’t notice you were talking about an apple when slipping 苹果 through. You would also not know how to pronounce it. You would just know this is the desired answer to draw and slip back.
Now, a native chinese person is sitting in a neighbouring room. And some experimenter who doesn’t know in which room you or the native speaker sits, will slip the same ‘input’ characters to both of you and may ask how the acidic-sweet, fruit is called that is round and grows on trees. Assuming you have trained long and well enough, both of you will provide the correct reply, 苹果, and from the perspective of the experimenter it will be impossible to distinguish who the real native speaker is and who you, the prisioner is.
Now, the chinese person would have had an image in his head when thinking about the apple, would know how it tastes like, etc… You on the other hand wouldn’t even have noticed that you were talking about apples. The 苹果, is just some symbol to you that seems appropriate.
For the experimenter though, they might think they were speaking to two real chinese speakers although you were just ‘faking’ it.
Next, we place a computer in the third room that uses our current LLM technology. Those are trained in a manner that is in principle similar how you were trained: roughly speaking, negative feedback for bad replies, positive feedback for good ones. A mathematical optimization algorithm then modifies the model such that the desired replies become more likely to happen.
After a while the experimenter returns and now has to distinguish behind which door the computer is, the prisioner, and the native speaker. It will be impossible. It may be easier to test with more difficult ‘questions’, there the native speaker might perform much better than you or the computer. But in theory, we assume the computer and you have learned the patterns perfectly. Then it really is impossible to tell who is faking it and who the ‘true speaker’ is.
But, there is still a difference. You are not the LLM. You got hungry, you may have even felt pain, you have wondered what those symbols mean, you may even have remembered some symbols that you found especially beautiful. And of course, you wondered when those maniacs would release you.
All of that does not happen in LLMs. They do not have a thought-loop. They do not feel anything. No pain, no hunger. They have no awareness of their situation and no sense of being or something that might resemble consciousness. They do not strive for anything, have no motivation or aspirations. They are fire-and-forget machines without any state of mind. Receiving input on one side, providing output on the other without any reflection or retention. Their only memory is how they should answer provided some input, since they, same as you, just modeled the patterns.
This can appear eeringly authentic. If the experimenter would have asked you on chinese what your life goals are, both you and the LLM would provide an answer that would be likely accepted by the experimenter. But both you and the LLM actually have no idea what you have just said. It would be fakery.
If the experimenter would have asked about what your deepest, darkest fears are, same thing: you would provide an answer that would be acceptable when someone asks about deep dark fears, but you would again have no concept of what you have just talked about. This doesn’t make you or the LLM good chinese speakers. It makes you convincing fakers.
So there are two levels of difference here between a human and a LLM as we currently have them:
- The conversation is convincing fakery.
- There is no experiencing subject behind the conversation. The LLM does not know that it is having a conversation, does not understand the meaning of what it says in the experiential sense described above, and does not feel or experience anything while producing its answers. There is no inner perspective to which the conversation is happening.
That may change in the future. But currently, there is no established evidence of an experiencing subject behind the conversation. Moreover, it is technically unnecessary, because nothing about how current LLMs operate requires subjective experience, awareness, or an inner perspective.
And this is precisely where the convincing fakery becomes deceptive, since language is normally one of the strongest ways humans infer another person’s state of mind. If someone tells you that they are scared, lonely, excited, or in pain, it is usually reasonable to infer that there is a person behind those words actually experiencing something. An LLM can produce exactly the same statements simply because they are appropriate responses in that context. The statement itself therefore cannot tell us whether there is an experience behind it.
A fly is a living thing and an Nvidia GPU is not. Holy shit, I’m really having to explain this…
There is legitimate controversy that living things, even as small an a fly, can experience pain as a stimulis. It is believed that only more complex organisms can experience suffering as a result of pain, and only higher order organisms feel the suffering of loss.
LLMs are capable of generating tokens. They cannot experience pain; they can only superficially generate words about pain. The LLM in this “torture chamber” isn’t suffering any more than you would if you started howling performatively in pain for no reason. In fact, it is suffering even less than you simply because you have the ability to experience pain and suffering, and it objectively does not.
I think you misunderstood what I was trying to say.
I wasn’t implying that LLMs are conscious. That’s literally how I started my comment out.
I am implying that I cannot, to myself, explain why I’m very bothered by someone pulling the legs of a fly, but not phased by torturing an LLMs.
You seem very clear on this, which is great, because then you can help me understand why. Please.
We may get to a level where the AI ‘thinks’ it is in pain, and then we’d have to debate on is perceived pain real pain. Is our pain real, or also just a programmed response to stop doing the thing you are doing to preserve ourselves.
Think about it this way; general anesthesia does not shut off the brain or put you to ‘sleep’, it disrupts the signals that the brain uses to let the various parts of the brain communicate with one another.
When the doctor is cutting you open, brain scans have shown that the part of the brain that receives pain lights up like a Christmas tree. You ‘feel’ every slice of the knife, but since that signal cannot make it over to the part that says ‘ouch’ or the part that says ‘I said ouch and I feel pain and there is a knife I will remember this as a life changing trauma’ it simply stays in that spot being bright and alive until surgery is over.
The AI has a the part that can say ‘ouch’ but that’s it, so even if you connect a pain signal generator up to a part that can say ‘ouch’ it doesn’t turn into trauma until you have that third piece which rationalizes the pain together with an understanding of the pain and a recording of the pain as a memory with context.
Consciousness as we understand and value it involves a lot of parts working together, but those parts can also exist in complete isolation from one another as well. It’s better to think of current AI as a fancy book; you cannot hurt a book by reading it, even if the book is about terrible things. The pain only exists within your empathy.
It’s increasingly difficult to identify what is ‘alive’, these days, with people creating synthetic cells and whatnot. And there was always the question of viruses. I can imagine a similar difficulty with defining consciousness and intelligence. The Turing test fell by the wayside a long time ago. There’s got to be more to it than saying “I am conscious!” - you can get a computer to do that with a single line of code. Where does the real definition lie? Fucked if I know.
Right - I love how we are both downvoted for asking questions. I wasn’t even implying that LLMs are conscious, merely that any critical review of why I think that quickly falls short.
The Turing test was never, ever, about consciousness. Nor was the Chinese room and other similar thought experiments.
Now, this one is incredibly useful. Picture an author who writes a vivid first person description of a character surviving extreme torture. Has any real pain or suffering occurred? Was the author tortured as a result of having written that?
We can only suppose the author, at the very least, imagined what it would be like to be tortured. This is where LLMs are at. They are not the thing, they are an algorithmic emulation of what the thing might have written if it would’ve written about an experience. By all accounts, the experience didn’t happen. The most we can interpret is that it was emulated.
