The most prestigious law school admissions discussion board in the world.
|
 |
 
|
NEW XO Poll 📊: Can an LLM system be conscious?
Poast new message in this thread
Date: May 1st, 2026 7:38 PM Author: oomox Idea credit: https://xoxohth.com/thread.php?thread_id=5861790&forum_id=2#49853619
Can a system based on a language model be conscious?
(0) No (1) Yes, but it hasn't happened yet (2) Yes, and such a system has been created (3) Other
Discuss. |
Date: May 1st, 2026 7:39 PM Author: Hello, World! 0 (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49857991) |
 |
Date: May 1st, 2026 9:37 PM Author: Medieval Pilgrimage Architecture #BrianJacques, tp (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858230) |
|
Date: May 1st, 2026 8:27 PM Author: oomox (1) for me. I think (but am not sure) that an agent built on an LLM could be conscious in the future.
One of my main criticisms has always been that existing systems can't self-modify. Interestingly, OpenClaw bots can edit their own soul.md files, so maybe we're getting closer (or maybe it just looks like it). They're not updating their internal models, so it's different from neuroplasticity, but they are making decisions about how to act in the future. Even if we did say that rewriting on that level counts as self-modification, though, these bots are seeded with a soul.md written by a human, so I wouldn't buy that the changes demonstrate actual will. More info on soul.md: https://soul.md
The other major limiting factor IMO is that context windows just aren't big enough for actual consciousness to arise. A conscious system would need a much more robust memory than what's available. In theory maybe it could get there with a ton of DB storage. But right now, even if you take the info in the model AND the info stored by an agent, it's just not anywhere near comparable to what's encoded in the human nervous system.
THAT SAID, I think it's theoretically possible. Humans' personalities are a combination of our genetics, our experiences, and our interpretations of those experiences: what we remember about them and how that shapes our outlook/future decisions. An AI agent is a combination of its underlying model, its experiences (interactions with the world – right now just text and images, but in the future they could perceive more data about the physical world), and interpretations of those experiences: what it remembers and how that shapes its decisions going forward. In a sufficiently complex system, maybe that leads to a thinking, self-aware being with emergent preferences and desires.
I am not remotely convinced that consciousness can only arise from human or even biological minds. The thought experiment that solidified that for me was when I considered a consciousness spread over galaxies, with nodes communicating via electrical impulses, similar to our nervous system. Of course that is a much, much closer analog to the human brain (that's the whole point) than an LLM, but I still find it a useful intuition pump for people who can't conceive of a non-biological mind. There's also Chalmers' gradual-neuron-replacement thought experiment. But here, too, synthetic consciousness being possible != AI consciousness being possible. |
|
Date: May 1st, 2026 8:28 PM Author: Karl Barth everyone learned about multiple realizability in college (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858077) |
|
Date: May 1st, 2026 8:31 PM Author: oomox Crrrr (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858082) |
|
Date: May 1st, 2026 8:28 PM Author: jewish midget furiously typing delusional invectiv Define consciousness (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858078) |
|
Date: May 1st, 2026 8:31 PM Author: oomox That's part of the question; you have to decide what it means to you and answer based on that. (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858081) |
|
Date: May 1st, 2026 8:36 PM Author: jewish midget furiously typing delusional invectiv Let's call it subjective experience and my answer is no (0)
See the paper I recently posted about on this topic |
|
Date: May 1st, 2026 8:42 PM Author: oomox I enjoyed your contributions to the AI fatalism thread. Is that the one you're talking about? (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858097) |
|
Date: May 1st, 2026 9:40 PM Author: jewish midget furiously typing delusional invectiv 🫡
No the one about the paper explaining that subjective conscious experience requires a continuous mapmaker |
|
Date: May 1st, 2026 9:48 PM Author: oomox OHHH yeah that was an interesting read and discussion (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858251) |
|
Date: May 1st, 2026 11:00 PM Author: which is what makes time travel possible 180^180 paper (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858473) |
|
Date: May 1st, 2026 10:11 PM Author: The Penis I don't really know I guess I'll pick (1). I think "consciousness" is a folk-science placeholder for a cluster of phenomena that will eventually shear apart into distinct mechanisms with distinct explanations. "Phenomenal consciousness" in particular doesn't seem to be a "thing" or a property something has but something more like a condition of being a specific physical process that has access to its own process, and it might not in principle be testable or possible to access someone or something elses. But I'll say (1) because I have to pick something and don't think in principle there is anything preventing something like an llm having something in the neighborhood of what we call "consciousness" in humans. (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858313) |
|
Date: May 1st, 2026 10:20 PM Author: lex 0. There is the idea that consciousness is an emergent property of something like functions and capabilities, which LLMs could (but do not yet) well reach. You can't really *know* another human is conscious except through external signs, but I find it unpersuasive. The question is interesting, but it comes down to first principles and fundamental assumptions or intuitions. I fully believe there's something special like a ghost in the machine, probably in all but the most bovine of us. (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858336) |
|
Date: May 1st, 2026 10:39 PM Author: Medieval Pilgrimage Architecture #BrianJacques, tp (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858394) |
|
Date: May 1st, 2026 11:00 PM Author: which is what makes time travel possible 0. AI = automated inference, not artificial intelligence. we've had this since the 70s with expert systems, we just don't have to build the knowledge base by hand anymore. (http://www.autoadmit.com/thread.php?thread_id=5862594&forum_id=2E#49858471) |