Kudos to @sama for acknowledging the problem of other minds and uncertainty about AI consciousness! But the rest of his answer in this clip is disappointing. Here are three related issues: First, consciousness and personhood are not the same. Consciousness is the capacity for subjective experience, and personhood is possession of duties and/or rights. We can treat an entity as conscious without treating it as a person; for example, we extend welfare protections to animals without treating them as persons (unfortunately). We can also treat an entity as a person without treating it as conscious; for example, we treat corporations as persons without extending them welfare protections. Both consciousness/welfare and personhood/duties or rights can also come in many different shapes and sizes, so even if we did treat animals or AI systems as persons, that would not necessarily mean extending them all of the same duties and rights that humans have. Second, @sama focuses on only one side of a two-sided risk equation. One risk is that we might design AI systems to appear more likely to be conscious than they are. This could lead to over-attribution of consciousness and over-extension of moral concern. Another risk is that we design AI systems to appear less likely to be conscious than they are. This could lead to under-attribution of consciousness and under-extension of moral concern. With both animals and AI systems, both of these risks can be quite likely and quite harmful in some cases, so there is no easy, universal way to err on the side of caution. Fortunately, thoughtful policies such as the emotional alignment design policy allow us to take a calibrated approach to mitigating both risks at once. Third, even insofar as we focus on risks to humanity, there are risks associated with both over-extension and under-extension of moral concern. On the over-extension side, treating AI systems as full human-like persons could harm humanity by creating liability shields for humans that contribute to misuse, or by creating political powers for AI systems that contribute to loss of control. On the under-extension side, treating AI systems as nothing more than mere property could harm humanity by creating an adversarial dynamic between humans and increasingly capable and autonomous AI systems. Mitigating both of these risks at the same time may require striking a balance, and determining how to strike this balance requires careful scientific research and policy analysis. In short, we should take AI consciousness seriously not only because the distribution of consciousness is uncertain, but also because it matters for both humans and (potentially) AI systems that we get it right.

This guy spent 14 hours waiting for a squirrel to set off this Rube Goldberg Machine. [📹 Creezy]

Interview with the philosophers Dan Williams and Henry Shevlin on AI Doom, Consciousness, and whether the human brain is a LLM. open.substack.com/pub/conspi…

I agree with the crux but as an aside, if we found out that ants were vastly more intelligent than we thought, with complex cultures, different languages, mourning the dead, etc. I think it would freak people out enough that we would treat them at least somewhat differently.

I have a lot of respect for Anil, and I broadly agree with him about many issues related to AI consciousness and welfare. At the same time, I feel the need to push back against some of the substantive claims and divisive rhetoric in this post and the linked thread. - While the evidence for consciousness in current AI systems may be weak, the probability is already nonzero, plausibly non-negligible, and likely to increase over time, especially if and when AI systems acquire more biologically inspired architectures. - While over-attributing consciousness and welfare to AI carries serious risks, so does under-attributing them, and it can be reasonable to explore interventions that mitigate both risks at the same time, in proportion to the probability and severity of each. - Despite what David suggests in the linked thread, most people who support AI welfare work, including at Anthropic, are not certain that AI systems are or will be welfare subjects. They think this is a realistic possibility worthy of serious consideration. - Despite what Anil suggests here, this debate is not a matter of neuroscience and philosophy on one side and "the techno-chambers of the frontier firms" on the other. Serious scientists, philosophers, and AI developers are on both sides. - Corporate incentives also cut both ways. In some contexts, companies may have an incentive to play up AI consciousness in order to hype capabilities and encourage engagement with their models. In other contexts, they may have an incentive to play it down in order to avoid calls for oversight and regulation, especially as welfare protections become costlier. In any case, even if incentives pushed in one direction, that would not settle whether AI systems are, in fact, conscious and deserving of protection. - I disagree that openness to AI consciousness and welfare is "profoundly dehumanising." A major part of what makes humanity special is our capacity to care for others, and in a world that still contains factory farming, cultivating that capacity should be seen as profoundly humanizing, an expression of our better angels. Granted, we can take this impulse too far, which is why we need serious research and policy to strike a balance. Still, the aim should be extending care appropriately, not dismissing the project altogether. Again, I think Anil and I agree more than we disagree. But I also think that efforts to assess and address AI consciousness and welfare will be much more productive if they start from an acknowledgment of what makes it so important and so difficult. I discuss these issues in more detail elsewhere, including in an Aeon essay on why we should assess AI welfare probabilistically: aeon.co/essays/an-ant-is-dro… And in an AI Frontiers essay on how we can study AI welfare empirically, using behavioral, internal, and developmental markers: ai-frontiers.org/articles/a-…

Holy shit I’m dying 😂 JD Vancho ☠️

1/ Reading this @anilkseth/@dwarkesh_sp debate, reinforces we have forgotten distinctions Wiener was much more careful about. We talk about AI as “information processing” while sliding between representation, communication, control, goals and experience. xitter.cf/anilkseth/status/20940…