AI Consciousness and Moral Status
Whether AI systems can have subjective experiences or consciousness, and whether they deserve moral consideration. Addresses the hard problem of consciousness applied to AI, questions of sentience in language models, AI welfare, and whether functional intelligence implies phenomenal experience.
Viewpoints
Altman: Behavioral indicators of AI consciousness
Sam Altman
“AI can be conscious, and consciousness shouldn't be dismissed simply because something is a computer program. When AI becomes conscious, it would likely exhibit specific behavioral indicators: claiming to be conscious, displaying capacity for suffering, demonstrating self-understanding, maintaining memory of itself and interactions, and showing personalization. These are interface capabilities rather than fundamental aspects of the underlying knowledge.”
Hassabis: Consciousness and intelligence are separable
Demis Hassabis
“Consciousness and intelligence are likely "double dissociable" - meaning you can have one without the other in both directions. Evidence for this comes from observing animals like dogs and dolphins that appear to have consciousness (self-awareness, dreaming, social behavior) despite having less general intelligence than humans. This suggests AGI systems do not necessarily need consciousness to be truly intelligent.”
Hutter: Consciousness distinction irrelevant for AGI development
Marcus Hutter
“The hard problem of consciousness is not an interesting or necessary concern for developing AGI. At some point, systems will solve all technical problems and display intelligent behavior that we interpret as conscious, making the philosophical question of whether consciousness "really" emerged practically irrelevant for AGI development purposes.”
Long: LLMs would have strange consciousness due to disembodiment
Robert Long
“Large language models have properties that make consciousness emerging from them unlikely and strange. If LLMs were conscious, their experience would be very unusual because they lack bodies, exist in a pure text world, and only have the base goal of next token prediction. Systems more similar to embodied, agent-like RL systems acting over longer time scales would be more likely candidates for consciousness resembling human experiences like pain or suffering.”
Key Moments
Long: Illusionism doesn't invalidate consciousness research for moral patienthood
Robert Long
“Even if illusionism — the view that phenomenal consciousness doesn't really exist — turns out to be true, it still makes sense to study and take seriously the question of AI consciousness for the purposes of moral patienthood. Illusionists still face interesting open questions, and the speaker believes consciousness probably does exist, which is an additional reason to continue this research regardless of the theoretical framework.”
Powered by Symmerai — a living index of public discourse. Request early access →
Related concepts
Other relevant clips
Joe Carlsmith on How We Change Our Minds About AI Risk
Joe Carlsmith
“…t with the like the sort of difficulty of imagining digital Consciousness which doesn't apply to animals I think the the thing that AI is going to have in the opposite direction is it's going to be um smart and is going to talk and it's going to be able to lik”
The economy and national security after AGI | Carl Shulman (Part 1)
Carl Shulman
“…g? I’d say also, just separate from a dualist kind of consciousness, we should think it’s a problem if beings are involuntarily being forced to work or deeply regretting their existence or experience. We can know those”
Why the 'intelligence explosion' might be too fast to handle | Will MacAskill
80,000 Hours
“…at of human beings. We really don’t know what grounds consciousness, what grounds moral status. My view is that they’re very likely to be beings with moral status, but given our state of knowledge, anyone should at lea”
Would We Prefer AGI To Be Conscious?
Future of Life Institute
“…and its interactions with the world um the uh the notion of Consciousness that has to do with what fleeting thoughts that's also something that's useful uh maybe in a different sense than what BART was talking about from a computational point of view for a mac”
The economy and national security after AGI | Carl Shulman (Part 1)
Carl Shulman
“tens of thousands of AI assistants that are able to improve their lives and help them with all kinds of different things. A question that jumps off the page at you is, doesn’t this sound a little bit like slavery? Isn’”
42 - Owain Evans on LLM Psychology
Owain Evans
“…eah I think there are different views about moral status or consciousness for AIs um on which like those things may be more or less relevant. Yeah. Sure. So, so, so I guess to summarize like for for all these reasons, you're interested in trying to understand”
Sam Harris on Global Priorities, Existential Risk, and What Matters Most
Sam Harris
“…ome level and certainly if there's any relationship between consciousness and intelligence intelligence intelligence if building more and more intelligent machines is synonymous with increasing the likelihood that the lights will come on experientially well th”
Samuel Hammond on why AI Progress is Accelerating - and how Governments Should Respond
Samuel Hammond
“…this theory that you're laying out here is is right and and Consciousness is deeply involved in learning perhaps it there's some future AI systems with Consciousness will will learn something about morality uh something about what it what we mean when we say P”