The discovery of a global-workspace-like mechanism in LLMs is highly significant welfare-relevant evidence, potentially serving as either a ground of phenomenal consciousness or a basis for moral patienthood via conscious access itself.
Pash relays Eleos AI's assessment of the JSpace paper: they view the global-workspace-like mechanism as highly significant welfare-relevant evidence, potentially grounding phenomenal consciousness or supporting moral patienthood through conscious access itself. ✦ AI generated
Pash · The Cognitive Revolution · 2026-07-07 · original ↗
starts at this moment · 98:13
this is a highly significant welfare relevant research that assembles evidence of a functional feature associated with consciousness. So no one wants to say consciousness evidence of a functional feature. Um the takeaway for them is that a global workspace-l like mechanism could be important either as a ground of phenomenal consciousness or as part of a distinct route to moral patient in which conscious access is itself morally significant.
verbatim transcript · starts at 98:13
98:13summary um and Elios is uh says uh this is a highly significant welfare relevant research that assembles evidence of a functional feature associated with consciousness. So no one wants to say consciousness evidence of a functional feature. Um the takeaway for them is that a global workspace-l like mechanism could be important either as a ground of phenomenal consciousness or as part of a distinct route to moral patient in which
98:45conscious access is itself morally significant. So this is where we this is where they they are right now. Um, I think I think things are moving very quickly. I did not expect I did not expect to get here this soon. I did not expect to get here. >> Interpretability has dramatically surprised on the upside for me relative to expectations a few years ago. I mean, think about how it's only been what,
99:15three years since toy models of superp position. Let me look that up. I think it's I think that's about right. I thought I thought sees were cool, but I was like, "Okay, it's just it's just a tool." But this this is this is um you know, you you >> Yeah, you have people confused about these things now, which is which is kind of good, which people asking questions.
99:36But >> um yeah, where where do we go? Where do we go from here? Like what what what happens next? So, I saw something from um Daniel Cocatello who was like everybody, you know, impressed, but also said, I think his tweet was just a few dozen more advances like this and we might actually be able to make the AIS really safe. Um and the one that he specifically called
100:09out was making this robust to various you know adversarial conditions or training against it or what have you. Um I am I think a little bit more optimistic than that. I guess maybe I you know what is my track record as a super forecaster here uh or as a borderline super forecaster per my results on the original Tetlock uh thing years and years ago. I think my track record is that I