What is your mechanistic model of self awareness that yields this conclusion?
> It's a tool
Does your model suggest that tools can't have self awareness?
(*) Even if we hack around this and just do the usual trick of simply laundering statefulness to a higher level, in this case the context window being fed in, I fail to identify (**) a representation of its own state in these bodies of text that it'd be meticulously maintaining. I further fail to identify how it could be hidden or maintained, considering I control like half of it. The best you could ascribe it is a meticulous maintenance of a persona the user is talking to, but then that doesn't necessarily represent the model's internal state, the same way my own words here aren't doing so either. Difference being, I actually have one (I'm "on-line").
You'll sometimes catch models mixing up who's who and how many who-s there even are for example.
(**) I did wish for something hidden though, so maybe it's just concealed? The same way people can encode a lot more of their emotional and mental state than normal into text if they read and write a lot of it, I'm aware of research that suggested the same for LLMs, albeit I cannot cite it. Maybe those phrasing signatures are just alien to me and will never pop out. Either way, I'd expect researchers to stumble upon this during interpretability studies, and either they haven't, they have but it wasn't popsci adopted, or they're keeping awfully tight lipped about it. If you know of anything like this, your turn now, would be happy to learn.
I do wonder how reasonable it is to expect e.g. a single maintained identity though. Maybe it isn't?
(*) Another way to hack around this of course is to just precompute some internal "self-awareness states" and hop around between them. Probably the closest to what the models are actually "doing".
> a hidden representation of self that is continually tended to
This sounds like a personality? They act like they have one of those. It may be an illusion, and even if it isn't an illusion it is unlikely to be anything like the source (us), but they act like it.
> I further fail to identify how it could be hidden or maintained, considering I control like half of it.
Indeed you control everything about a local model, and much of the context of even a remote model. But the state of activations and circuits in SotA AI is hidden in similar ways to those of synapses in your head: difficult to decipher even with probes monitoring the signals directly, and often not emitted at the normal output.
> The best you could ascribe it is a meticulous maintenance of a persona the user is talking to, but then that doesn't necessarily represent the model's internal state, the same way my own words here aren't doing so either. Difference being, I actually have one (I'm "on-line").
While we can be confident that LLMs make up personas etc., it is insufficient to go from "that doesn't necessarily represent the model's internal state" to "therefore it doesn't have one".
> You'll sometimes catch models mixing up who's who and how many who-s there even are for example.
I've, unfortunately, also experienced this with humans. Perhaps they were losing their self-awareness at the time? I do wonder if old-age dementia does that by the end, though the person in question didn't ever get diagnosed with that.
> If you know of anything like this, your turn now, would be happy to learn.
Do you mean like these, or something else?
• https://researchportal.hkust.edu.hk/en/publications/decoding...
I don't see why an LLM could not have a sense of identity or personality while it's evaluating a specific prompt, or even change self awareness while evaluating a prompt since many outputs model a back and forth conversation. My point is that without a mechanistic model of what "self awareness" means, we have no way of truly evaluating such questions, we're just hand waving vague intuitions about what it could mean.
A language model (or an image model or whatever) cannot even be sentient, and I think sentience is a prerequisite for awareness.
Even if we express a lot of our subjective experience with words, the language is just a symbolic representation of those experiences. The qualia themselves, even those that are quite abstract, are rooted in our physical presence and evolution.
You can't have an understanding of what hunger or physical pain feel like if you have no need for food or a sensory capacity for feeling pain. You can't understand what loneliness or pride at an achievement mean if you don't have a neural network wired to value social connection or status. We value connection because we're social animals that have needed each other for survival.
Even the more abstract of our subjective experiences are in some way rooted in our physical evolution.
I see no reason to believe that a neural network built entirely based on the symbolic level of language could have the features needed for the subjective experience itself.
AI awareness might actually be more believable if that awareness manifested itself in an entirely different way than in humans. But if we assume awareness because outputs resemble what we consider meaningful as humans, yet the neural network has had no inputs or evolution that could form the actual basis of human-like experience, I think we're seeing something that isn't actually there.
There is no objective evidence of qualia. All evidence of qualia are vocal or other expressions of belief in qualia. Perceptions clearly exist and are observable, subjective experience and qualia, not so much.
> I see no reason to believe that a neural network built entirely based on the symbolic level of language could have the features needed for the subjective experience itself.
If your objection is to models based on "symbolic level of language" which you think lack semantic understanding of, say, trees, you should ask yourself how our brain, based on physics which also lacks any semantic category for trees, can somehow develop a semantic understanding of trees. All of these appeals to differences with the brain never seem to acknowledge that fundamentally, the brain has the same explanatory gap with physics.
> But if we assume awareness because outputs resemble what we consider meaningful as humans, yet the neural network has had no inputs or evolution that could form the actual basis of human-like experience
This assumes a lot. It seems very possible to me that intelligence inherently develops a map of natural categories (natural kinds), and language naturally develops around such categorical understanding. Semantics are then fundamentally the network of associations between categories, eg. there is no fundamental difference between symbols and semantics, and the latter cam be inferred from the former, and that's exactly what LLMs do, and why the semantic maps between different languages are so similar and how they can translate between languages.