what if we can abstract, and model the neurons as spikes + control mechanisms + ion concentration densities + randomization, and non-neuronal effects similarly? But in order to simulate the brain using a full physics sim, you need QM, say to figure out the randomization bit, even though pseudo-randomness is just as good from a “how smart?” and personality perspective.

What?? This question is preposterous, of course the brain is quantum; it relies on proteins whose structure is determined by its constituent electrons energy levels which are themselves quantised. I have no stake and I'm not going to bet either way but this should resolve YES immediately imo.

All molecular phenomena require quantum mechanics to function as it does. I think a criteria that is much more in the right spirit is could you emulate the computations performed (not the exact underlying physics) with a classical computer, or do you need a quantum computer with it’s unique gates and operations?

There's things that aren't quite right with this criteria

@RemNi Like necessary vs sufficient things

As the editor of Jeffrey Satinover's

"The Quantum Brain: The Search for Freedom and the Next Generation of Man," I have to say yes.

I'm very confused by this definition. There are countless possible reasons why a classical computer might fail to simulate something that a quantum computer successfully simulates without this proving that quantum effects are required for a successful simulation. A simple programming error is the most obvious, or perhaps the hardware of a classical computer simply heats up too much or accumulates too many floating point errors, or perhaps the quantum computer necessarily simulates the atomic level rather than merely the molecular and this happens to be significant for some reason.

I also don't find "resembling human behavior" to be a very rigorous standard at all. What if it turns out that quantum effects are required for all multi-cellular organisms to function, albeit not all bacteria? What about if it's only all mammals? What if it's only humans and chimps? What if it's only required for humans, but also only required for them to have complex dreams while they are asleep? What if the human brain doesn't need quantum effects to function, except during its initial development in the womb? What if the classical simulation works flawlessly, but fails to accurately simulate a brain seizure, or brain death? What if the classical simulation is indistinguishable from the quantum simulation in almost every way, except the classical human brain requires significantly more energy to the point that a full classical human would not survive? What if the classical simulation works flawlessly, but only if the connected sensors in the eye (or some other organ) are simulated with quantum effects? What if the classical simulation only works if the light entering the eye is simulated with quantum effects, otherwise the person is blind? What if both the classical simulation and quantum simulation flawlessly produce "human behavior", but the exact same starting conditions result in two wildly different personalities, or morals, or mental disorders?

I feel like I could continue asking such questions indefinitely, but have restrained myself to merely 10.

I'm genuinely amazed that none of the prior comments brought up similar issues, instead preferring to focus on defining the line between classical and quantum. However one distinguishes between classical and quantum effects, all of these issues remain so long as there's any distinction at all.

To clarify, you are asking whether you need a quantum computer to efficiently simulate a brain, right? (Technically: brain simulation has BQP complexity, but not P.) So, if the brain has some quantum effects, but they aren't computationally relevant, it will resolve to NO, right?

If this understanding is correct, the current probability is unbelievably high. We have a plausible classical models of how brain might work in neural networks, and not even a vaguest idea of how quantum computing could be involved in the brain.

@OlegEterevsky My personal belief is around 5%. But this is one of those questions where it is not clear what will be the resolution source, and I find it quite likely that even the resolution might end up opinion based, so I am not betting it down.

Even among scientists, there are groups that have strong and opposite opinions, so the full consensus seems far. (And it is unlikely we will emulate a brain upload any time soon).

@IsaacKing Can you confirm that efficient classical simulation of functional brain would resolve this NO?

@Irigi Once we understand brain mechanism well enough to simulate it (and I'm hoping we are not too far from that moment), this could be resolved based on whether the brain could be efficiently simulated on a classical computer or not. That's pretty unambiguous.

Here's a somewhat relevant question that I wrote today:

To clarify, you are asking whether you need a quantum computer to efficiently simulate a brain, right?

Hmm, this seems like a reasonable definition, yeah. Basically I'm trying to get at whatever people mean when they talk about something being "quantum" in popular science articles.

Seems like there should be a distinction between something can be simulated classically and something cannot function without quantum effects. For example, is the vision system considered part of the brain for the purposes of the question?

@bashmaester Quantum mechanics can be simulated classically, only not with so good algorithmic complexity.

Does "then that isn't sufficient to resolve this YES" mean it resolves NO or does it mean it resolves N/A?

@ArmandodiMatteo Depends on what the actual answer is.

@IsaacKing what do you mean exactly?

Out of the four possibilities
1. a bacterium would function in a classical simulation, and so would a brain
2. a bacterium would function in a classical simulation, but a brain wouldn't
3. a bacterium wouldn't function in a classical simulation, so neither would a brain
(and in principle 4. a bacterium wouldn't function in a classical simulation, but a brain would -- though the chances of that are negligible, unless you use different standards for "functioning" and for "classical simulation" in the two cases)

1 (or 4) would resolve NO, and 2 would resolve YES.
You're saying that 3 wouldn't resolve YES, but would it resolve NO, or would it resolve N/A?

I strongly suspect that quantum effects provide the brain with the ability to process information not sequentially or in parallel, but simultaneously. Subjectively, this looks like a so-called gestalt. I think I will live to see the day when this is proven.

@Shalun do you have any ideas of how that could work?

@CodeandSolder I personally think quantum entanglement is relevant. A more detailed answer should be in the book The Relativistic Brain: How it works and why it cannot be simulated by a Turing machine. But I still haven’t read it, because there is no Russian translation, and I procrastinate reading in English.

@Shalun quantum entanglement can be simulated by a Turing machine in exponential time, so the book is wrong or irrelevant if you want to know about quantum effects rather than some kind of spooky uncomputable new physics.

@AMS from an Amazon review of the book:

In the authors’ example, a protein (as a tiny [Oracle] machine) finds its optimal 3-D configuration – an intractable computational problem – in an instant by following the laws of physics in the analog domain.

I count four or five different serious errors in this single sentence:

• Proteins are not oracle machines; they run on Turing-computable physics

• They do not always find the same "optimal" configuration (ever heard of a prion?)

• Finding the lowest energy configuration is Turing computable, and only intractable in the weaker sense of maybe being NP-hard (I'm not certain of the class)

• They don't fold in an instant, the timescale is quite significant on a molecular scale

• (Bonus) The laws of physics may be analog but thermal systems are noisy, so the analog/digital distinction is not important for computational considerations here.

@AMS and the only reason the ones we use are that robust in finding the desired configuration is that using ones that don't is a huge evolutionary disadvantage lol

As others have alluded to, "the classical laws of physics" is extremely ambiguous -- there are many different ways to use non-quantum physics to approximate quantum physics, including most of chemistry. Most "quantum effects" can be treated classically in one way or another. Every argument about whether a phenomenon is a "quantum effect" devolves into muddy semantics, in my experience. Photons are a quantum effect from the perspective of classical electromagnetism (Maxwell's field theory). Electromagnetic fields are a quantum effect from the perspective of classical particle physics (as applied to photons).

As a clean dividing line, there are some quantum phenomena that can't be efficiently simulated classically, notably quantum computation. If the brain is "a quantum computer" in this sense, then it can't be simulated classically without an exponential-in-brain-size slowdown. This isn't completely rigorous (to a computer scientist) because brains have a constant size, but I expect it to (eventually) be extremely clear that the brain can either be simulated in something like "polynomial time" by a classical computer, or it can't.

So I would propose: the brain is quantum if it cannot be simulated in polynomial(brain_size) time by a classical computer.

@AMS (Penrose seems to be making a stronger claim, but I think it would be totally fair to call the brain quantum even if it were "merely" a BQP machine.)