Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The New AI Consciousness Paper

astralcodexten.com

Post reply on HN
Print view View original

201–210 of 313 posts

Re: The New AI Consciousness Paper

#201

Post by mlinsey »

Animats wrote: post #186

The most insightful statement is at the end: "But consciousness still feels like philosophy with a deadline: a famously intractable academic problem poised to suddenly develop real-world implications." The recurrence issue is useful. It's possible to build LLM systems with no recurrence at all. Each session starts from the ground state. That's a typical commercial chatbot. Such stateless systems are denied a stream o…

I predict as soon as it is possible to give the LLMs states, we will do so everywhere.

The fact that current agents are blank slates at the start of each session is one of the biggest reasons they fall short at lots of real-world tasks today - they forget human feedback as soon as it falls out of the context window, they don't really learn from experience, they need whole directories of markdown files describing a repository to not forget the shape of the API they wrote yesterday and hallucinate a different API instead. As soon as we can give these systems real memory, they'll get it.

Re: The New AI Consciousness Paper

#202

Post by lukifer »

dsadfjasdf wrote: post #194
andai wrote: post #193

My summary of this thread so far: - We can't even prove/disprove humans are consciousness - Yes but we assume they are because very bad things happen when we don't - Okay but we can extend that to other beings. See: factory farming (~80B caged animals per year). - The best we can hope for is reasoning by analogy. "If human (mind) shaped, why not conscious?" This paper is basically taking that to its logical conclusio…

If something convinces you that it's conscious, then it effectively is. that's the only rule

If it is the case that consciousness can emerge from inert matter, I do wonder if the way it pays for itself evolutionarily, is by creating viral social signals.

A simpler animal could have a purely physiological, non-subjective experience of pain or fear: predator chasing === heart rate goes up and run run run, without "experiencing" fear.

For a social species, it may be the case that subjectivity carries a cooperative advantage: that if I can experience pain, fear, love, etc, it makes the signaling of my peers all the more salient, inspiring me to act and cooperate more effectively, than if those same signals were merely mechanistic, or "+/- X utility points" in my neural net. (Or perhaps rather than tribal peers, it emerges first from nurturing in K-selected species: that an infant than can experience hunger commands more nurturing, and a mother that can empathize via her own subjectivity offers more nurturing, in a reinforcing feedback loop.)

Some overlap with Trivers' "Folly of Fools": if we fool ourselves, we can more effectively fool others. Perhaps sufficiently advanced self-deception is indistinguishable from "consciousness"? :)

Re: The New AI Consciousness Paper

#203

Post by kjkjadksj »

andai wrote: post #193

My summary of this thread so far: - We can't even prove/disprove humans are consciousness - Yes but we assume they are because very bad things happen when we don't - Okay but we can extend that to other beings. See: factory farming (~80B caged animals per year). - The best we can hope for is reasoning by analogy. "If human (mind) shaped, why not conscious?" This paper is basically taking that to its logical conclusio…

The factory farming argument is a little tired. I'd rather be killed by an airgun over what nature intended: slowly eaten alive by a pack of wolves from the anus first.

Re: The New AI Consciousness Paper

#204

Post by yannyu »

criddell wrote: post #155
hermitShell wrote: post #141

Earlier quoted context omitted.

I think this is an excellent point. I believe the possibility of 'computing' a conscious mind is proportional to the capability of computing a meaningful reality for it to exist in. So you are begging the question: Is it possible to compute a textual, or pure symbolic reality that is complex enough for consciousness to arise within it? Let's assume yes again. Finally the theory leads us back to engineering. We can at…

Is anybody working on learning? My layman's understanding of AI in the pre-transformers world was centered on learning and the ability to take in new information, put it in context with what I already know, and generate new insights and understanding. Could there be a future where the AI machine is in a robot that I can have in my home and show it how to pull weeds in my garden, vacuum my floor, wash my dishes, and a…

This is where the robotics industry wants to go. Generalist robots that have an intelligence capable of learning through observation without retraining (in the ML sense). Whether and when we'll get there is another question entirely.

Re: The New AI Consciousness Paper

#205

Post by fizx »

ctoth wrote: post #99
yannyu wrote: post #21

Let’s make an ironman assumption: maybe consciousness could arise entirely within a textual universe. No embodiment, no sensors, no physical grounding. Just patterns, symbols, and feedback loops inside a linguistic world. If that’s possible in principle, what would it look like? What would it require? The missing variable in most debates is environmental coherence. Any conscious agent, textual or physical, has to inh…

> A conscious textual agent would need something like a unified narrative environment with real feedback: symbols that maintain identity over time, a stable substrate where “being someone” is definable, the ability to form and test a hypothesis, and experience the consequences. So like a Claude Code session? The code persists as symbols with stable identity. The tests provide direct feedback. Claude tracks what it wr…

I tend to look at consciousness as a spectrum. And when we reduce it to a binary (is it conscious?), we're actually asking whether it meets some minimum threshold of whatever the smallest creature you have empathy for is.

So yeah, Claude Code is more conscious than raw GPT. And both probably less than my dog.

Re: The New AI Consciousness Paper

#206

Post by BobbyJo »

andai wrote: post #193

My summary of this thread so far: - We can't even prove/disprove humans are consciousness - Yes but we assume they are because very bad things happen when we don't - Okay but we can extend that to other beings. See: factory farming (~80B caged animals per year). - The best we can hope for is reasoning by analogy. "If human (mind) shaped, why not conscious?" This paper is basically taking that to its logical conclusio…

> My question here is... why do people do bad things to the Sims? If people accepted solipsism ("only I am conscious"), would they start treating other people as badly as they do in The Sims? Is that what we're already doing with AIs?

A simple answer is consequences. How you treat sims won't affect how you are treated, by other people or the legal system.

Re: The New AI Consciousness Paper

#207

Post by ajs808 »

Animats wrote: post #186

The most insightful statement is at the end: "But consciousness still feels like philosophy with a deadline: a famously intractable academic problem poised to suddenly develop real-world implications." The recurrence issue is useful. It's possible to build LLM systems with no recurrence at all. Each session starts from the ground state. That's a typical commercial chatbot. Such stateless systems are denied a stream o…

https://arxiv.org/pdf/2304.03442

Re: The New AI Consciousness Paper

#208

Post by BriggyDwiggs42 »

kjkjadksj wrote: post #203
andai wrote: post #193

My summary of this thread so far: - We can't even prove/disprove humans are consciousness - Yes but we assume they are because very bad things happen when we don't - Okay but we can extend that to other beings. See: factory farming (~80B caged animals per year). - The best we can hope for is reasoning by analogy. "If human (mind) shaped, why not conscious?" This paper is basically taking that to its logical conclusio…

The factory farming argument is a little tired. I'd rather be killed by an airgun over what nature intended: slowly eaten alive by a pack of wolves from the anus first.

That’s ridiculous though. A normal life for an animal involves lots of hardship, but also pleasure. Factory farms are 24/7 torture for the entire life of the animal. It’s like being born in hell.

Re: The New AI Consciousness Paper

#209

Post by jswelker »

LLMs have made me feel like consciousness is actually a pretty banal epiphenomenon rather than something deep and esoteric and spiritual. Rather than LLMs lifting machines up to a humanlike level, it has cheapened the human mind to something mechanical and probabilistic.

I still think LLMs suck, but by extension it highlights how much _we_ suck. The big advantages we have at this point are much greater persistence of state, a physical body, and much better established institutions for holding us responsible when we screw up. Not the best of moats.

Re: The New AI Consciousness Paper

#210

Post by andai »

lukifer wrote: post #202
dsadfjasdf wrote: post #194

Earlier quoted context omitted.

If something convinces you that it's conscious, then it effectively is. that's the only rule

If it is the case that consciousness can emerge from inert matter, I do wonder if the way it pays for itself evolutionarily, is by creating viral social signals. A simpler animal could have a purely physiological, non-subjective experience of pain or fear: predator chasing === heart rate goes up and run run run, without "experiencing" fear. For a social species, it may be the case that subjectivity carries a cooperat…

>If it is the case that consciousness can emerge from inert matter, I do wonder if the way it pays for itself evolutionarily, is by creating viral social signals.

The idea of what selection pressure produces consciousness is very interesting.

Their behavior being equivalent, what's the difference between a human and a p-zombie? By definition, they get the same inputs, they produce the same outputs (in terms of behavior, survival, offspring). Evolution wouldn't care, right?

Or maybe consciousness is required for some types of (more efficient) computation? Maybe the p-zombie has to burn more calories to get the same result?

Maybe consciousness is one of those weird energy-saving exploits you only find after billions of years in a genetic algorithm.

Post reply on HN