Let's talk orchestrated objective reduction!
My special interest of late has been something called https://en.wikipedia.org/wiki/Orchestrated_objective_reduction the TL;DR is:
- One of, if not the, most missing piece of quantum mechanics is answering the question "what is measurement"? You've probably heard of things like the double slit experiment which lead to weird things like quantum erasure where one can seemingly cause a photon to retroactively determine which path to take. Spooky stuff! However, these experiments all follow the basic idea of "when something is entangled, it follows probabilistic rules defined by the schrodinger equation, then a 'measurement' happens, and the entanglement 'collapses' and only a single, 'real' value is well defined"
- Roger Penrose, a nobel prize winning physicist, has, since the 1980s been arguing that because entanglement implies that a particle exists in two places at once prior to measure that this places gravitational pressure on the fabric of spacetime in two places at once and that measurement is a gravitational event where spacetime "heals" itself by collapsing the wave form, and making it so the particle is finally only in one place.
Penrose further expanded this, to enormous controversy, that consciousess itself is a measurement event. He wrote a book, "The Emperor's New Mind" then a follow-up "Shadows of the Mind", neither of which I've read, but have had summarized to further develop these arguments. - This was, for lack of a better term, a crackpot theory. There wasn't anything testable or falsifiable so it was brushed aside. Crucially, to the point it gets its own paragraph,
The overwhelming opinion of the physics community, to this day, believes that quantum coherence is not possible in "wet, warm, noisy" environments like the brain.
It is, to this day, believed that, the quantum world is a thing that happens only at extremely small scales, and that's why quantum computers all start with the assumption of cooling the material to near absolute zero with as few additional perturbations as possible. - However, there were 2 findings I find extremely motivating to combat this assertion. First, leaves. Leaves are quantum objects and photosynthesis is too efficient to be explained by classical mechanics alone: https://berkeleysciencereview.com/article/2021/11/30/plants-do-the-wave/ Second, birds. Birds that use the magnetosphere for orientation do so by becoming quantumly coherent with the ions in the magnetophere using a protein in their eyes so they literally see the earth's magnetic field using the blue cones of their eyes https://mitadmissions.org/blogs/entry/birds-are-real-and-so-is-quantum-physics/
- Enter Stuart Hameroff. Hameroff was a working anesthesiology for some 20 odd years. He read The Emperor's New Mind and reached out to Penrose. He, of course, was very intimately aware of the deep biological processes that make the different between a conscious, aware, thinking human being, and a piece of meat that can be safely operated on. He believed that https://en.wikipedia.org/wiki/Microtubule structures in the neurons were responsible for consciousness.
After further research, they both began to believe that these microtubule structures in human neurons were capable of what others believed were impossible.
Quantum coherence in a wet, warm, noisy environment. - This was still crackpottery until quite literally (IMO) last year. A group of biological researchers showed experimentally that the exact networks of tryptophan microtubule structures in neurons do exhibit super radiance (the same kind of quantum coherence leaves show) https://pubs.acs.org/doi/10.1021/acs.jpcb.3c07936
To me, this is the most exciting piece of science I think I've seen in my life. The implication is that human (and other) consciousnesses are literally a byproduct of our ability to maintain and calculate quantum states in a wet, warm, noisy environment. It feels like a genuine push for us to finally move past the age of information, past the age of computation, and into the age of consciousness.
Another aspect of this is I really, deeply, believe that the substrate necessary for AGI is either necessarily biological, or, at the very least, can only be done efficiently in a biological substrate. Notably, a human brain takes 20 watts to exhibit generalized intelligence. No nuclear reactors running data centers, just a Twix bar. This last bit I honestly leave mostly as a point of discussion, because there's an enormous amount of interesting implications and avenues thereof.
What y'all Tildeans make of this? Anyone else been thinking about this kind of stuff?
I am deeply sceptical. I did my PhD in quantum physics, and I think people too desperately search for meaning and consciousness in any little gap of our understanding. I don't know why consciousness and free will are so important for people to fit into the physical models, despite all difficulties associated with doing so.
A few scattered thoughts:
On measurements in QM:
You can write extensive books on measurements: talk about the observer becoming part of the superposition; stipulate about Everett's many-worlds and ponder your fellow yous in different worlds; or calculate classical world line equations. It's all valid, and all useless. The reason why most physicists follow the "shut up and calculate" school of quantum mechanics (i.e., the interpretation doesn't matter, QM is a mathematical approach that predicts things well, and that's all that you should care about) is the complete lack of objective value in any other approach. What matters in physics is falsifiable predictions; is being able to show experiments that support your theory (or even better, disprove your fellow physicists theories). None of the different ruminations on the deeper philosophy of QM meet that bar; they are all elaborate ways of coloring in the gaps between what we can verify. A fun philosophy topic for elderly physicists used to being the smartest person in the room, maybe.
On the overall topic of... Quantum consciousness?:
To me, consciousness and free will are ill-defined as a physics-concept. Are we talking something that is not deterministically predictable by our models (potentially for the wave function and thereby stochastic after a measurement, but with at perfectly knowable probability distribution)? Then to our knowledge, it doesn't exist. Are we talking something that is part of the models? Then we don't need consciousness and free will in there; it already has a causal reason. Or are we talking about something that occurs on larger scales? Then it's an emergent phenomenon; super cool stuff but you can have those without any quantumness just fine, any computer can manage that.
I suppose the idea is to suggest that the presence of quantum-mechanical effects of some kind in organic matter supports that consciousness could be somehow related to quantum mechanics, but does it really? To me, it just means that there are degrees of freedom in organic systems that are reasonably isolated from their environment and therefore able to form some coherence, which is not that special - take e.g. NV centers in diamond, which manage much more quantumness just fine at room temperature. I don't see how it makes the connection of quantum mech to consciousness any stronger. Every atom inside your neuron is already fully quantum. All physical interactions are. Your neuron is fundamentally a quantum object; just not a very coherent one. It's rare to see it on the macroscopical level; but that's because we never truly observe and conceptualize a closed system, and open systems are messy and hard.
Fundamentally, to me, consciousness, free will, and the quantum mechanical measurement are not relevant topics in the physics sense of "how does this help me understand the laws of nature through theory and experiment". They are philosophy. Philosophers are, of course, allowed to find value in anything they wish, but they should then call it philosophy, not physics; not until it makes a falsifiable prediction, which I strongly doubt any of these questions will because of the sheer improbability of understanding a system of even the size of a single neuron quantum-mechanically.
On Penrose: the guy is really smart, no doubt. But I will say that there is a tendency for older physicists to play at philosophy, with a confidence in their reasoning that only physicists can have. Life and Consciousness are big topics to many of them; take Schrödingers work, for example. But it is important to keep on mind that being a brilliant physicist does not make you a brilliant philosopher; it does not even mean that you will remain a brilliant physicist. Neither do accolades and a prestigious professorship; look only to Brian Josephson as a sad example.
I am a physicist, with rather interdisciplinary research, but with some connections to areas that Orch-OR seeks to cover (more on the biology and computation side, less on the quantum side and nothing on the philosophy or spirituality side). I would echo your points here, which are perhaps put more sensitively than I would have written them.
I think there is a tendency in the scientific community to be respectfully quiet to the general public, when the situation isn't too egregious, about flawed research of aging prominent researchers. There is a human fear and sadness for declining faculties, made perhaps more terrifying by being a community where those faculties are everything. Within the community, I have heard nothing but skepticism, at best, for Orch-OR; a professor in a related field once expressed his view to me privately along the lines of I don't have a deep knowledge of many of the topics involved, but for the areas I do, the work seems so heavily flawed in even their basic ways of thinking about those areas that I'm very skeptical of all of it. I have also heard the suggestion that the Penrose-Hameroff relationship could be seen as that of a crackpot preying on an older, prominent scholar's wandering to push his own ideas: unlike Penrose, it should be pointed out that he has no background in physics or even science generally (he was a medical professor), and has involvement in areas of clear pseudoscientic quantum mysticism, for example, being in What the Bleep Do We Know!?; he is apparently now working on a feature-length animated film to promote his theories. I should add to all of this: I have a great respect for much of Penrose's work, and I think that many people who see Orch-OR as nonsense do too.
There is the distinct feeling that the project, potentially for both of them, comes out of a certain desire to show something 'special', or perhaps more aptly, 'sacred' about the human mind that distinguishes it from something seen as 'mundane', then searching to find anything that, through enough leaps, might suggest something sacred. That view is disappointingly dismissive of both 'mere classical computation', and of the many fascinating ways that computation can take place in biological systems.
I'd also note: with physics in particular, there seems to be a tendency for the field to capture the imagination of the general public, but in ways that are often very misleading as to what research in the field entails, what current research and questions are, and the prominence and establishment of theories and topics. We can see that throughout this thread. Physics as presented to the general public can often end up being more about vague philosophizing, and attempts at analogies to explain theories that are actually hard math, where the analogies end up being seen by some in the public as having meaning in themselves. It tends to heavily focus on philosophical interpretations of quantum mechanics (not really a major research question, as you point out), cosmology, high-energy particle theory, and string theory (usually without clarifying that string theory's position and reputation in physics is arguably very dubious).
Wonderfully put. I guess it is very human to want to be more than the sum of inconceivably many deterministic little events, and to want to identify something beyond that that makes us who we are.
To me, its somewhat the opposite: There is such beauty and wonder in how mine own complex self emerges out of the little stochastic determinism of elementary particles underneath... I am more than content in being "merely" an emergent phenomenon so complex, we could never hope to truly understand it :)
On your last note: I cannot fault anyone for wanting to understand the inner workings of the world more; I cannot fault people for reading pop-sci explanations of topics like quantum mech to understand more, as that is what is approachable to them; so maybe it is naive to complain about people putting too much weight on the analogies in that content, which is all they know on the subject. I truly wish it was possible to communicate the depth of the math behind the theories without requiring 4-6 years of university level math first.
Speaking of free will and quantum mechanics, I recently came across the Conway and Kochen Free Will Theorem:
I left the field of physics after my undergrad and have been happily at play in the fields of chemistry and material science ever since. Also, my quantum was terrible, so I never felt comfortable with bras and kets. Maybe that's why the theorem appeals to me.
I agree with all of this, and honestly, it's exactly why I left that last paragraph in. Penrose's ideas of "proto consciousness" arising out of the collapse of the wave function are fun, and his idea of super position causing gravitational wells in two distinct point in spacetime simultaneously causing a pressure on the fabric of spacetime are similarly fun, but there's nothing rigorous there.
All that aside, as someone that can clearly talk to talk, how do you feel about the possibility that the modern paradigm of chasing quantum computation in exclusively absolute zero conditions may be missing the point because the concept of coherent quantum states in "warm, wet, noisy" biology has become effectively dogmatic?
That's the bit that is testable and is having results popping up more and more that I do find exceptionally exciting.
Assuming that Godel's incompleteness theory holds, and that there are non-logically true statements in physical reality, that implies that mathematics and formalized systems are fundamentally incapable of a full expression of the world around us. From that angle, I really am getting more convinced that the future of quantum computation should not be focused on the classical idea that qubits can/should/will follow a similar Moore's law curve, simply because the fundamental substrate of silicon (with doping, the Majorana chips out of Microsoft are in fact very very cool) that they are working with is, at best, very inefficient.
I actually began typing that it may be impossible, because true classical computation assumes the falsehood of digital information, but now that I think more, modern quantum computation seemingly doesn't have the similar trappings of acting like 0 and 1 are real actual objects in physical reality, rather than analog signals that we can play pretend with, since they (to my layman knowlege) are working on the more fundamental wavey substrate the quantum field theory is founded upon.
Sorry that last paragraph lost steam, but I am curious. Since leaves are definitely super radiant, isn't it at least a bit exciting that the dogmatic view "brains are too loud to be quantum" could be straight up wrong? My thoughts immediately jump to all the pitfalls of transformer based AI systems and how they very well could be fighting in a classical world and losing some 20-30% efficiency in the process? My "scary tagline" is, maybe AGI will just run most efficiently off vats of organoid intelligences (brains in jars)? Any thoughts I can pick from you out of that?
I happened to work on building a quantum computer for a while, and we did choose an ultra-cold system for it, so I am quite convinced that it is the right approach. Mostly, this is just a question of timescales - as a rule of thumb, colder systems means less interactions with the environment means more time until decoherence. At hot temperatures, most timescales are blazing fast compared even to our best computers; so its hard to really do computation on a hot quantum system - it might be enough for a quantum effect to be visible (i.e. your examples in the OP), but for quantum computation we want to run algorithms, meaning we want to perform a bunch of very precise operations in succession. Colder systems just mean you have the time to control that algorithm with a regular computer :)
Beyond that remark, I am not quite clear on what you are actually asking with regard to quantum computation. I dont see a benefit in trying to do QC in a warm system - almost trivially, taking whatever system that is and cooling it down should perform better, as lifetimes increase etc. Beyond that, the actual system and temperature always only matter in relation to its inherent timescales - we worked with Rydberg Qubits, which are way colder than superconducting Qubits (but so far still worse, maybe); but they are also much more easy to cool, so it kind of cancels out? Lots of considerations that go into selecting a system.
As for the other questions... I mean no offense here, but those last paragraphs read very scattered between topics and come of more manic to me than anything else. You are jumping from math, to QC, to Moore's law (far from applicable to QC so far), to something about falsehood of digital information (?), quantum field theory, then to quantum effects in wet matter again, and then to transformer based AI systems all of a sudden...
All of those topics are worthy of study, and one could spend a lifetime on any one of them. Connecting them together is beyond our current understanding of many of them; and more than any human could achieve, even with the requisite decade+ of studying math and theoretical physics to start to understand some of them.
I dont know what background you have, but I will carefully posit that any theory by literally any human, living or dead, that claims to bring all these things together is of very questionable scientific merit. There are so many unanswered questions on the way that must be answered first; there is no point in speculating that far out into the unknown.
Coming at this from a cognitive science perspective, I'd argue that we lack a robust and widely agreed upon definition of "consciousness" to begin with, which makes it difficult to implicate in any sorts of theories involving it.
Further -- absent a citation -- I'd assume that the folks who're asserting the impossibility of quantum computation being done at a wide scale in the human brain are doing so because quantum entanglement (the natural phenomenon which we exploit with quantum logic gates to perform calculations) decoheres at higher energy levels (e.g. room temperature) or when exposed to other particles smashing around (e.g. in a cell, bombarded by all sorts of intracellular goo).
Finally, since I'm not a biologist, I can't handwave away your observation that the macro-scale impact of quantum effects are all over the animal kingdom if only we were to look hard enough. Fair! I'd note, however, that they're everywhere in the manmade world, too, as indicated by the double slit experiment (example from Wikipedia). Or even the fact that electrons quantum tunnel through the nanoscale transistors in your CPU, complicating their design! The trick is that quantum effects != quantum computation.
Now for some sidebars:
That's why Hameroff's contributions are so crucial to the theory. We don't have a good definition of consciousness but I'd argue we have a good definition of unconsciousness. When you go under for surgery, the job of the anesthesiologist is to turn off your consciousness carefully and keep it off so you are just raw meat to be worked on. Similarly, sleep is arguably a state of reduced consciousness and sleep medicine has categorized various levels of consciousness with "deep" sleep being the furthest thing from being awake and aware that we get to in a given night (REM is, IIRC a level above this).
This was my point about leaves. They exhibit super radiance, AKA coherent quantum entanglement with photons, in the harsh environment of the outdoors. Without this coherent super radiance, leaves photosynthesis can only reach something like 70-75 energy efficiency when it's been observed that leaves reach in the high 90s% of energy efficiency. One literally needs the mathematics and science of quantum mechanics to fully explain how leaves work.
From there I make the jump that Hameroff is making that the hydrophobic microtubules in neurons should exhibit similarly quantum properties, and, as of 2024, this was proven experimentally, that UV EM waves exhibit super radiance in the microtubules of neurons at room temperatures.
The consciousness stuff is fun to ponder, but this is a scientific revelation to me. The dogma of "no quantum coherence because noise" is likely correct, however, it's seeming more and more that science has dramatically underestimated our own mind's capability to create micro pockets of silence to maintain coherence in.
The leap that I similarly struggle with is "if you turn off that quantum state machine in the neurons, you turn off the consciousness". As far as I've read, this is still in "reasonable hunch" territory for Hameroff in particular. My greatest shoutout to Penrose in all this is, if you listen to interviews, he is emphatic that he does not understand the biology, and he doesn't attempt to make more than a layman's approach to explaining what's going on. Hameroff is the superhero here when it comes to getting actual, experimental, falsifiable evidence out.
Fair cop!
I counter with https://www.youtube.com/watch?v=aY985qzn7oI Luis Alvarez is the perfect counter point that sometimes, scientists are just raw geniuses that go around solving a huge range of wildly disparate problems, but they are the exception to the rule. Mostly, as Angela notes, as you age, you suffer neurodegernation which is a tragic truth of the human condition. The example she gives of this "gracefully aging into crackpottery" is Francis Perey, his grandaughter wrote an article about his slow descent into losing his coherence and right state of mind around "solving" Bell's Theorem here https://www.theatlantic.com/science/archive/2018/11/science-full-mavericks-like-my-grandfather-was-his-physics-theory-right/574573/
Which is to say, I'd suggest you listen to Penrose and keep the 4 checks of a crackpot theory in mind:
YES!! This is precisely why I'm actively working to shift my career away from these cold, quiet, falsehoods of classical information theory and computation into the exciting world of organoid intelligence. You have hit the nail on the head that today you can literally load up a Python API and start reading and writing to them using Multi Electrode Arrays. Right now the problem for scaling these organoid systems is, interesting enough, not the actual neuron growth, but the vasculature to get oxygen to the whole organoid so it can grow to more than a pinkie nail (or less?) of cells.
Your point on the "someday" thing for quantum computers is a part of this excitement for me. I've actively investigated going into quantum computation but man, it just feels like a field that's been stalling out for 10-15 years now. The Hype People came in saying "it's a computer, of course qubits will follow moore's law" only to find, no, maintaining coherent quantum systems doesn't scale nearly as easily as scale coherent information systems. That's why this biological intelligence stuff has me so, so very jazzed, is that from where I'm sitting it has a very real possibility of flipping the entire paradigm of quantum computation on its head!
I don’t know enough biology to read this paper. However, I think you’re overestimating how energy-inefficient LLM inference is? For one thing, they’re not running all the time, just when generating a response.
Coding agents do run continuously for quite a while, so that does seems like relatively heavy usage.
For what it’s worth, ChatGPT’s estimate for a coding agent is 15-55 watts, not including your laptop, which might be 10-15 watts, along with 25 watts for an external monitor.
I don't have a lot to add, but was working on sci-fi concept for little bit which ran with the premise: "what if orchestrated objective reduction is real and the key to strong AI?" I found it difficult to communicate the ideas clearly and things just ended up reduced down to a MacGuffin anyways. (That all probably says more about my ability than anything else, though.) The exact mechanics likely don't for storytelling purposes, but that's what makes it easy to easy to be a little handwavy and entertain the idea.
I would still like to go back to that concept and world at some point. The more recent headlines keep it somewhat front of mind.
While the theory can sound like a series of logical steps, overall I tend to lean toward determinism and am still very skeptical about it.
I would meme on it if anybody I know would get the joke.
LOL that last sentence is perfectly. I must admit, I have a lot of trouble subscribing to some of Penrose's deeper and wilder theories on the consciousness part of things. The bit I really am, more or less, conviced, is that we're all quantum machines.
That bit just makes sense to me becuase quantum field theory really is just, the best we've got, in terms of explaining physically explaining what's going on. To me, biology is built on physics is built on math. The true battlefield that evolution and biology must logically be waging war on is not classical but quantum simply because quantum reality is reality.
That's a big part of why "leaves exhibit super radiance quite convincingly" is such a big part of it to me.
To your actually point though, Penrose's ideas on the collapse itself being the raw material of consciousness and that things like photons decohering is a "proto consciousness" is, well, yeah, like you said, sci fi stuff.
As a continuation, had you considered how these concepts may play out in a TTRPG setting? Having physical processes be expressed in game rules could be a more instructive approach? Spitballing, I'm a long time DM
What he did is convincingly prove a single cell organism is still fully, surprisingly conscious with good memory and just about every single reaction to stimuli that we have. With that one simple zen-like kill shot he deleted brains (and therefore all cognitive science degree-holders) from any future discussions of consciousness. This is once again a physics problem. That's in fact tremendous progress, no more brain-related bullshit on the table hiding the real answer to this question. :)
Just recently Stuart appeared on Danny Jones (2.5 hours) and did a phenomenal job of explaining what he's talking about. It is handily the best conversation about consciousness I've ever seen. He brought the receipts too, citations abound.
I'm not sure he's right about his and Roger's fundamental premise - that quantum collapse is itself a burst of conscious experience. It's an intriguing way of thinking about the universe, that's for sure, but we've got to pin it down with experiments. That's going to be tricky but he has got the problem in a bit of a box now.
The implications of this if he's right are a fatal kill shot for sapient artificial general intelligence, because if he's right, no configuration of silicon is ever going to simulate even a single cell effectively, it's beyond our data storage capabilities. Kiss mind uploading goodbye along with the rest of the singularity crew's pipe dreams. Won't have to worry about AI waking up because it hasn't got the kind of equipment necessary in hardware to do the waking up - and we can prove it, so no rights for the robots after all. Electrons can't play these games.
Did someone say the phenomenology of Quantum Mechanics?
That Penrose and Hameroff are on the hunt for consciousness in macroscopic quantum effects in the h&man body is REAL interesting. I'll certainly have to read more about that. Do you have some good intro material other than the papwr, the wikipedia article and Penroses own books?
Accepting quantum stuff in the wet, warm and noisy doesn't seem too far fetched to me. But I wonder how that keads to consciousness? What is consciousness by the Penrose Hameroff definition, and how does quantum calculations lead to that?
Also, a minor thing, I'm curious what you think the age consciousness looks like? I'm all for a good next step, and I think it's interesting to hear what others believe the next step is.
To me the biggest issue is what the quantum computation inside of neurons is supposed to explain. As far as I know the non-quantum models of neurons work just fine, so there's not much in the behavior of single neurons that need explaining. Then we know how neurons communicate between each other and that also doesn't need much explanation, and it also seems dubious quantum states could be transmitted via electrical spikes & neurotransmitter release. If the answer is consciousness, well mere quantum computation in microtubule just doesn't explain consciousness at all so that doesn't help either.