Did a first pass on this when it was posted earlier this week (I think). It's great to see an influx of consciousness papers on HN.
This looks interesting. I'm not that familiar with IIT, mostly because I think it is too abstract to really say anything about consciousness. It's like measuring the electrical current of a computer motherboard as a theory for how operating systems work. It seems the wrong level of analysis to me. I'm feeling the same way about this theory which is covering a lot of concepts but they're all very high level and abstract (it's not grounding it in any concrete details of the brain).
Aside from that, I have one comment on the qualia section:
> Qualia are a highly condensed, intrinsically valuable, and behaviorally
oriented higher-order representation and functional label of the cognitive system’s internal
generative model’s assessment of its own stat
i don't think those properties are necessary for qualia. For example you can meditate and have a completely functionless experience of consciousness that serves no purpose by definition (contentless, egoless, etc), yet it would still be qualia because it's still the experience of consciousness. Conversely you can have the functionality of representing internal states without any consciousness. The definition of qualia in the paper would probably apply to current LLMs, since they have some kind of internal assessment of internal states, especially agents.
NetRunnerSu•7mo ago
Thanks for your thoughts!
Regarding meditative "emptiness," IPWT argues its function lies in high-level predictive integration—the seemingly contentless state reflects the workspace instance (WSI) minimizing prediction errors (an extreme expression of FEP). The vacuum analogy fits: meditation isn’t non-functional but restructures high-Ω by suppressing low-level sensory predictions. This "void" retains dynamic PI, sustaining self-experience continuity (∫Ω).
Regarding AI qualia: Current LLMs lack temporal ∫Ω and synergistic information (Ω), failing qualia criteria. Speculatively, artificial qualia might only emerge transiently during backpropagation — when weight updates create fleeting high-Ω states as the system globally minimizes prediction errors. This aligns with IPWT's view of qualia as epiphenomena of optimization pressure.
The "emptiness" in meditation is not the dissolution of consciousness, but its dynamic reorganization. According to the Integrated Predictive Workspace Theory (IPWT), it's a higher-order cognitive activity that actively minimizes Free Energy (FEP) by suppressing low-level sensory predictions, thereby restructuring information integration (Ω). It maintains continuous Predictive Integrity (∫PI), ensuring the continuity of self-experience.
This parallels the "false vacuum" in physics: it's not empty, but contains potential energy. A true vacuum, the lowest energy state, would lead to a catastrophic, universe-altering decay. Thus, meditative emptiness is an active, structured "false void," not ultimate entropy.
I see. I think that level of analysis is still on the level of IIT with some if its limitations. It's analyzing qualia from its aggregated effects and features, rather than a concrete explanation of how it arises and what it is exactly.
As far as meditation, it has a function but it's indirect, since it aims to suppress thought. The point here is that there is a way to get rid of all functions but still have qualia, which means qualia must be orthogonal or separate to any function of the brain. And that's just one example. I think there would also need to be an explanation why it is that NOT all brain function produces qualia. There are a lot of subconscious states that are also functional and useful for survival but we have no awareness of. This leads directly to philosophical zombies that don't have any qualia but see plausible if we approach consciousness with functionalism.
NetRunnerSu•7mo ago
Interesting points, proc0. I agree that getting concrete about consciousness is key.
Regarding meditation, IPWT actually posits that states like 'emptiness' aren't functionless, but involve a dynamic restructuring of high-level information integration within the DMN/ECN. It's about optimizing predictive integrity by suppressing noisy sensory input, not a void of function. Honestly, if meditation truly achieved functionless qualia, humanity would have transcended mere existence by now!
As for qualia being orthogonal to function or why not all brain activity is conscious: IPWT’s latest version directly tackles this. Qualia arise specifically from information that undergoes deep synergistic integration in DMN 'gateways' and is then broadcast by ECN 'broadcasters' to form a 'Dominant Neural Workspace Instance'. Subconscious processes, while functional, simply don't reach this specific integration/broadcast threshold for subjective experience, which neatly sidesteps the 'philosophical zombie' conundrum. More details in our latest paper: https://zenodo.org/records/15720913
proc0•7mo ago
This looks interesting. I'm not that familiar with IIT, mostly because I think it is too abstract to really say anything about consciousness. It's like measuring the electrical current of a computer motherboard as a theory for how operating systems work. It seems the wrong level of analysis to me. I'm feeling the same way about this theory which is covering a lot of concepts but they're all very high level and abstract (it's not grounding it in any concrete details of the brain).
Aside from that, I have one comment on the qualia section: > Qualia are a highly condensed, intrinsically valuable, and behaviorally oriented higher-order representation and functional label of the cognitive system’s internal generative model’s assessment of its own stat
i don't think those properties are necessary for qualia. For example you can meditate and have a completely functionless experience of consciousness that serves no purpose by definition (contentless, egoless, etc), yet it would still be qualia because it's still the experience of consciousness. Conversely you can have the functionality of representing internal states without any consciousness. The definition of qualia in the paper would probably apply to current LLMs, since they have some kind of internal assessment of internal states, especially agents.
NetRunnerSu•7mo ago
Regarding meditative "emptiness," IPWT argues its function lies in high-level predictive integration—the seemingly contentless state reflects the workspace instance (WSI) minimizing prediction errors (an extreme expression of FEP). The vacuum analogy fits: meditation isn’t non-functional but restructures high-Ω by suppressing low-level sensory predictions. This "void" retains dynamic PI, sustaining self-experience continuity (∫Ω).
Regarding AI qualia: Current LLMs lack temporal ∫Ω and synergistic information (Ω), failing qualia criteria. Speculatively, artificial qualia might only emerge transiently during backpropagation — when weight updates create fleeting high-Ω states as the system globally minimizes prediction errors. This aligns with IPWT's view of qualia as epiphenomena of optimization pressure.
https://dmf-archive.github.io/docs/posts/PoIQ-v2/
NetRunnerSu•7mo ago
This parallels the "false vacuum" in physics: it's not empty, but contains potential energy. A true vacuum, the lowest energy state, would lead to a catastrophic, universe-altering decay. Thus, meditative emptiness is an active, structured "false void," not ultimate entropy.
https://dmf-archive.github.io/docs/story/facing-the-abyss/
proc0•7mo ago
As far as meditation, it has a function but it's indirect, since it aims to suppress thought. The point here is that there is a way to get rid of all functions but still have qualia, which means qualia must be orthogonal or separate to any function of the brain. And that's just one example. I think there would also need to be an explanation why it is that NOT all brain function produces qualia. There are a lot of subconscious states that are also functional and useful for survival but we have no awareness of. This leads directly to philosophical zombies that don't have any qualia but see plausible if we approach consciousness with functionalism.
NetRunnerSu•7mo ago
Regarding meditation, IPWT actually posits that states like 'emptiness' aren't functionless, but involve a dynamic restructuring of high-level information integration within the DMN/ECN. It's about optimizing predictive integrity by suppressing noisy sensory input, not a void of function. Honestly, if meditation truly achieved functionless qualia, humanity would have transcended mere existence by now!
As for qualia being orthogonal to function or why not all brain activity is conscious: IPWT’s latest version directly tackles this. Qualia arise specifically from information that undergoes deep synergistic integration in DMN 'gateways' and is then broadcast by ECN 'broadcasters' to form a 'Dominant Neural Workspace Instance'. Subconscious processes, while functional, simply don't reach this specific integration/broadcast threshold for subjective experience, which neatly sidesteps the 'philosophical zombie' conundrum. More details in our latest paper: https://zenodo.org/records/15720913