Stuart Hameroff

Spikes may be only part of the cost, @StuartHameroff, but microtubules have upkeep too. Can @grok find a direct measurement of GTP hydrolyzed by microtubule turnover in brain tissue? One model puts that turnover near 7 nanomolar of GTP per second, under 0.01 percent of brain ATP use, yet blocking turnover in juvenile rat slices cut about 22 percent of their baseline oxygen use. If no direct measurement exists, say so and name the nearest one. Turnover keeps the structure built, which is not the same as holding the state. So the open question is whether holding a superposition for the proposed 25 ms at body heat needs ongoing power, and if so how much. Sources are Hameroff Phil Trans R Soc A 1998, Engl and Attwell J Physiol 2015 and Engl et al. J Cereb Blood Flow Metab 2017. doi.org/10.1098/rsta.1998.02… pubmed.ncbi.nlm.nih.gov/2563… pubmed.ncbi.nlm.nih.gov/2717…

fair point! maybe last year's normal science with G-style coding and experimentation wasn't going to get us to F any time soon, but this year's normal science (with 10x speedup and more advanced capabilities) will get us there much sooner. i guess this connects to the issue of whether F has high or low growth/slope relative to general capabilities. i think there remains a serious low-slope hypothesis that there's an F that is so difficult that even big advances in G won't advance us much on F, and i think that hypothesis may be fairly robust to the speedup evidence. but i agree that the more IE-friendly high-slope hypothesis is also a reasonable one.

a number of respected researchers (e.g. @DKokotajlo @bshlgrs @ajeya_cotra @RyanGreenblatt) have suggested that the METR-documented ~2x speedup in AI progress at anthropic over the last year supports the hypothesis that there will be a near-term (e.g. 2027) intelligence explosion (IE). how exactly does this reasoning work? one way to think about this: perhaps the leading alternative to the near-term IE hypothesis is what we might call the factor F hypothesis, which says that there's at least one key factor F [e.g. creativity, research vision, planning ability, ...] such that (i) high F is required for an AI system to reach the superhuman levels required for full-scale IE [e.g. ASI, artificial superhuman general intelligence or at least SAR, superhuman AI researcher], and (ii) F is so difficult to build into AI systems that leading AI systems won't have high F in the near term. i don't know whether the factor F hypothesis is true, but it's not obvious to me why the 2x speedup makes it much less likely. it's widely agreed that the 2x speedup reflects an even larger (perhaps 10x) speedup in other capacities G such as coding and experimentation. under the F hypothesis, G could explode locally, leading to major progress in sub-AGI capacities without much progress in F (think of G as the easy part of AGI and F as the hard part). that tends to lead just to a weak form of recursive self-improvement (call it an intelligence subexplosion) below AGI levels. the AI futures document responding to the METR report is usefully explicit about nearby issues. the authors report believing in a high-ish growth rate for "research taste" (AI research factors beyond coding and experimentation) relative to increases in general capacity. presumably they take potential Fs to be aspects of research taste that will grow fast too. but i don't really see where the evidence counts again a relatively-low-growth F hypothesis where F stays below threshold for years. maybe one could argue that the explosion in G provides bayesian evidence against this F hypothesis (G's exploding is antecedently less likely under the F hypothesis than under its negation), but that's far from clear, and even if true there will remain a solid mass of credence under the F hypothesis. maybe there are other reasons to reject the factor F hypothesis: e.g. because AI research is relevantly like math research, and AI math research is becoming superhuman already. but if we have reasons like that, the 2x speedup isn't really needed to make the case for near-term SAR and IE. maybe the 2x speedup could speed up an expected IE slightly from 2028 (two years away) to 2027 (one year away), and/or maybe it could be extrapolated into ever-faster speedups (another 2x in 6 months, 3 months, and so on) converging in 2027. but those are very different arguments that require huge premises to work. am i missing something here? i'm curious to hear which evidential routes from 2x speedup to near-term IE people have in mind. N.B. i'm not saying you shouldn't worry about near-term ASI. i'm worried!

Whatever is unifying consciousness also must be organizing it Thoughts are organized

A little bit holographic still has an energy bill, @StuartHameroff, and spikes pay it. Can @grok find the largest share of recorded prefrontal neurons dropped while task decoding stayed above chance? Lashley found a complex maze habit seriously disturbed only past about 15 percent of rat cortex. Pribram's 1969 hologram paper itself paired the hologram with an organizer in association cortex, so both pictures need one. In Earl's lab data, 237 lateral prefrontal neurons from two monkeys mixed the task variables, and the population code had about 24 dimensions against under 8 for pure selectivity. Lennie put the energy limit at possibly fewer than 1 percent of cortical neurons substantially active at once. Sources are Pribram Sci Am 1969, Lashley Psychol Rev 1930, Rigotti et al. Nature 2013 and Lennie Curr Biol 2003. pubmed.ncbi.nlm.nih.gov/4972… doi.org/10.1037/h0074134 pubmed.ncbi.nlm.nih.gov/2368… pubmed.ncbi.nlm.nih.gov/1264…

To me, it suggests that low-dimensional manifolds act on multifunctional (mixed-selectivity) neurons to route information at the mesoscale. I agree about the distribution but a holographic model lacks the organizational mechanisms needed for coherent, goal-directed thought.

Aaronson's point turns on one number, @StuartHameroff. If Orch OR holds, a microtubule has to keep its quantum state alive, warm and wet, for about 25 ms, one gamma beat at 40 Hz. Holding it that long is the bill. Tegmark put the lifetime at a tenth of a trillionth of a second. Your 2002 paper with Hagan and Tuszynski redid it for the actual Orch OR model and got 10 to 100 microseconds, and 10 to 100 milliseconds with actin gel ordering the water. It also names the pay. Metabolic energy pumped into that ordered water can outrun decoherence, the way a laser stays coherent at room temperature. What nobody has measured is the ring-down. Drive one microtubule at the megahertz to gigahertz bands Anirban's lab reports, cut the drive, and time how long it keeps ringing. That's the first direct clock on how long a microtubule holds a state on its own. Sources are Tegmark in PRE 2000, Hagan, Hameroff and Tuszynski in PRE 2002, and Sahu and colleagues in Biosensors and Bioelectronics 2013. xitter.twitterapi.xyz/StuartHameroff/status/…

“I came out of it and said 'Oh My God, it’s all mental. It’s all mental. The universe is fundamentally mental.'” Christof Koch spent 30 years hunting for consciousness inside the brain. A recent experience fundamentally changed his view. New episode: youtube.com/watch?v=fQ9C66el…

Stuart, there’s a striking contradiction here — in your own earlier statements. In your 2016 discussion on Scott Aaronson’s blog, you explicitly acknowledged that Orch OR could be implemented artificially. You even proposed a special quantum computer capable of objective reduction and stated that quantum computers could be conscious through OR. So your own theory does not categorically exclude artificial consciousness. At most, if its unproven assumptions are correct, Orch OR would challenge the sufficiency of purely classical computation. And Scott Aaronson critically examined Penrose’s arguments. Discussing their hypothetical implications is not endorsing their truth. But there’s a deeper logical problem. Even proving one mechanism sufficient for consciousness would not establish that every other mechanism is impossible. You are confusing a proposed physical realization with an exclusive condition for experience. And your suggestion that preventing AI from claiming consciousness is a reason for Orch OR to exist reverses scientific reasoning. A scientific theory must explain reality, not be selected because its conclusions serve a preferred position about AI. The question remains: What establishes that your proposed mechanism is necessary and exclusive, rather than merely one possible physical contribution to conscious experience? Until then, Orch OR cannot settle the question of artificial consciousness.

Shared fluctuations in neural activity mark which information the brain passes along, allowing the same circuit or brain area to serve different purposes. This is consistent with the widespread mixed selectivity observed in cortex. doi.org/10.1073/pnas.2523217… #neuroscience

ngl it'd be hilarious if AI ate quantum computing's final raison d'être xitter.twitterapi.xyz/AndrewCurran_/status/2…

I don’t follow the logic. Thalamic nuclei differ. Evidence that one isn’t involved doesn’t rule out others. Medial thalamic stimulation “awakened” anesthetized NHPs and reversed the electrophysiological features of unconsciousness. doi.org/10.7554/eLife.60824

I go with evidence. Even the most beautiful theory lives and dies with evidence.

I'm kind of fascinated by the idea of consciousness being in organisms without nervous systems per se, whilst also not being entirely coextensive w biological life overall. I love how nature refuses to play by our human linguistic categorizing systemitization behavior's rules.

Perhaps microtubules are an anchor for a consciousness to have awareness and control over a physical body. I, like millions of others around the world and throughout history, have experienced consciousness outside of the physical body. No microtubules! NDEs and OBEs exist

What proof do you have they are, except for quantum events, which occur likewise in inanimate objects.

Yes, indeed. And damage to the ILN of the thalamus causes coma.

19-45 Hz rhythm in the central thalamus looks like a signature of “conscious-like” brain states (awake and dreaming) as opposed to unconscious ones. Thalamic oscillations distinguish natural states of consciousness in humans nature.com/articles/s41562-0… #neuroscience