Jeff Sebo

speaker

150 appearances 1 recordings 1 series first heard Jul 2026 last heard 10 Jul

Jeff Sebo’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months

Recordings per month over the last 12 months — 1 in all, peaking in Jul 2026 with 1.

Appearances

newest first · ▶ plays the moment

Hard Fork · Do Social Media Bans Work? + A Conversation About A.I. Consciousness + Tool Time · 10 Jul 2026 podcast

And so I hope that we can learn more in the coming years.

And then armed with that information, we of course also need to consider safety for humans and other animals, constraints on our resources,

and put that all together into policy decisions.

So this is going to be a very complicated process, and the outcome is probably going to be somewhere in between continue treating them purely like tools versus immediately give them human-like legal and political rights overnight.

It will probably be something in the middle of those extremes.

Yeah, expert, maybe a little bit of a crank.

I thought it was really significant, too.

Not necessarily proof of phenomenal consciousness in particular, not proof or even particularly strong evidence that the models are having feelings and emotions of a morally significant sort, but it is evidence of aspects of...

access consciousness.

It is evidence of aspects of what we call a global workspace.

And so global workspace theory is one of the leading scientific theories of consciousness.

And it basically posits that consciousness arises when a system has

a bunch of different modules, processing a bunch of different types of information, and then develops a central workspace where privileged information can be collected from the different modules and then processed and then broadcast back to those modules.

And this is a way of integrating and coordinating activity across all of the different activities happening in the brain.

And what is noteworthy about this study

is despite the fact Anthropic did not specifically design Claude in order to achieve this, they are discovering aspects of a global workspace-like space within the model.

And the fact that the models developed this kind of workspace in order to assist with their reasoning, in order to assist with their language production, is really striking and noteworthy and worth studying very carefully.

I would not endorse those specific criticisms.

I think that those are too skeptical about what this shows.

I think this is genuinely important research that is pointing to a new aspect of language models that gives us more insight into how they work and might be a

Showing 101–120 of 150 · page 6 of 8