Nobel laureate Roger Penrose is widely held to  be one of the most brilliant living physicists   for his wide-ranging work from black holes to  cosmology.

And then there’s his idea about how   consciousness is caused by quantum processes.

Most scientists have dismissed this as a cute   eccentricity—a guy like Roger gets to  have at least one crazy theory without   being demoted from the supersmartypants  club.

The most common argument for this   dismissal is that quantum effects can’t  survive long enough in an environment   as warm and chaotic as the brain.

Well,  a new study has revealed that Penrose’s   prime candidate molecule for this quantum  activity does indeed exhibit large scale   quantum activity.

So was Penrose right  after all?

Are you a quantum entity?

The story starts with Roger Penrose  as a young man attending a lecture on   the Gödel incompleteness theorems.

These are the  mathematical arguments by logician Kurt Gödel,   which prove the limits of proof.

The  first incompleteness theorem says that,   for any consistent formal mathematical system  capable of expressing basic arithmetic,   there are true statements in that system that  cannot be proven with the rules of that system.

Here’s an example of an unprovable statement:  This statement cannot be proved true in this   language.

Now imagine there’s an algorithm  that can search through all combinations   of expressions in the natural language in  question until it successfully found the   one that proved this statement true.

That  very act would render the statement false,   thereby showing that the system is inconsistent.

But if the algorithm couldn’t find such an   expression it would demonstrate that the  statement—which is that this statement   can’t be proved—is true but unprovable.

So  either the system is inconsistent or contains   true but unprovable statements—it's incomplete.

Natural languages are certainly inconsistent,   but mathematics is designed to be perfectly  self-consistent, therefore, mathematics   is incomplete.

And since any algorithmic  system is generally built on mathematics,   like classical computation, they too must  be incomplete, according to Kurt Gödel.

But Penrose argues that humans can  “prove” the unprovable.

For example,   mathematicians seem able to be pretty sure of  the truth value of certain conjectures without   any formal mathematical proof—and sometimes those  formal proofs may not exist.

According to Penrose,   that means our conscious process of knowing  does not come from a process limited by   Gödel incompleteness—it’s not algorithmic or  computational.

Penrose goes further to say   that this implies consciousness itself can’t  emerge from purely computational processes.

Similar ideas were articulated  by philosopher John Lucas,   so this whole “consciousness  isn’t classical computation   because Gödel incompleteness” thing  is called the Penrose-Lucas argument.

There are plenty of objections to this  argument—for example, that it’s a mistake   to equate our sense of knowing to a formal  proof.

After all, it’s possible to think you   know something but be wrong.

Or that even if  we do have super-Gödel reasoning abilities,   it’s a stretch to connect that to consciousness.

We’ll leave it to you to check out Penrose’s   books “The Emperor's New Mind” and “Shadow Of The  Mind” for further details if you’re interested.

OK, let’s get to the weird part.

If we’re not classical computers,   what are we?

Penrose argues that the  only place we might possibly find a   type of computation or information  processing or whatever that’s free   from the limits of Gödel incompleteness is  … maybe you guessed it, quantum mechanics.

Quantum mechanics describes the world of  the tiny.

Quantum objects like subatomic   particles or even molecules can have  very weird behaviors.

Their properties   like their location or speed tend not  to be well defined.

They can exist in   multiple states or even places at once—in  what we call a quantum superposition.

And   these fuzzy quantum properties can also be  correlated with other quantum objects in a   mysterious way called quantum entanglement.

This is all stuff we’ve described before.

It’s not hard to force a quantum object  to give up its quantum fuzziness and take   on well-defined values.

All you need to do  is measure it.

Measurement causes an object   in a superposition of multiple states to choose  one of those states randomly and eliminates any   entanglement.

We say that measurement “collapses  the wavefunction”--where the wavefunction is the   mathematical object that defines the distribution  of possible results of that measurement.

After   the measurement, one result is chosen and the  wavefunction collapses to that single value.

Exactly what causes this collapse or even   what precisely defines a measurement  is not known, even 100 years after the   discovery of quantum mechanics.

We call  this mystery the measurement problem.

One thing we know about wavefunction collapse  is that it appears to include a truly random   factor—measurement results are chosen by the  roll of the dice, albeit dice weighted by the   shape of the wavefunction.

So, if an event is  truly random then almost by definition there is   no algorithm that can perfectly determine  its outcome.

Therefore, Penrose argues,   the outcome of any computation or information  processing performed by a quantum system involving   wavefunction collapse is non-algorithmic, in the  sense that the outcome cannot be predicted within   the framework of any Gödel-esque mathematical  system—and therefore is not necessarily subject   to Gödel incompleteness.

Therefore maybe  conscious reasoning has a quantum component.

There are plenty of criticisms of this reason—for  example, quantum computation is still algorithmic   in a sense, even if not classically algorithmic.

Also, it’s not clear how the injection of a random   process frees one from the constraints  of Gödel in the way that Penrose needs.

Another common criticism is that Penrose has  fallen for the Holmsian fallacy.

To quote the   great Sherlock Holmes—When you have eliminated  all which is impossible, then whatever remains,   however improbable, must be the truth.

In fact  Penrose himself has quoted Holmes to justify   the reasoning that if consciousness can’t  be accounted for by things we understand,   then it must be accounted for by the other big  thing we don’t yet understand—quantum mechanics   and the measurement problem.

But that assumes  that Penrose has comprehensively eliminated,   or is even aware of, all other candidate  mysteries, which feels presumptuous.

On   the other hand, this is Roger Penrose we’re  talking about, so maybe he really has taken   a complete census of all mysteries and is  justified in Sherlocking consciousness.

A final objection, and the one that  will finally get us to the new result,   is that quantum behavior is generally only  observable in the most pristine conditions.

Typically for individual or smallish collections  of subatomic particles, and carefully isolated   from the environment.

Quantum states decay  extremely quickly unless in a vacuum and/or   at near absolute zero temperature.

And the  more particles involved in a quantum state,   the more easily the state gets destroyed.

That’s  why quantum computers are so hard to build.

The inside of the brain is far from  a pristine environment—it’s warm and   gooey and seems thoroughly macroscopic  and classical.

In order to do the sort   of quantum information processing that  the Penrose-Lucas argument demands,   a coherent quantum state needs to  be maintained for timescales much   longer and involve far more particles than  should be possible in our meat computers.

Penrose first published his idea in the Emperor’s  New Mind, but at the time had no idea how brains   might do quantum processing.

But then along came  Stuart Hameroff.

Hameroff is an anesthesiologist   who had developed a fascination with consciousness  from a young age, and subsequently with a   molecular structure inside cells called  microtubules, which he suspected may be involved.

Microtubules are these tiny tubes that play  many roles in every cell in your body.

They’re   a major part of Eukariot cell’s skeleton—the  cytoskeleton—stabilizing its shape.

They act   as conveyor belts, moving proteins around.

They  even play a key role pulling chromosomes apart   when a cell divides.

They’re constantly being  assembled and disassembled and adapt to what   the cell needs at any given moment.

A given  cell might have billions of microtubules.

Microtubules also have an extremely regular  structure—almost crystal-like.

They are made   of alternating tubulin molecules of two different  types.

Those molecules each have a polarity which   can point one way or another.

This structure  got Hameroff to thinking maybe microtubules   could in some way work as molecular computers,  or at least molecular information storage.

Add   to this two more hints: microtubules are  more abundant and differently structured   in neurons than any other cell.

Also, there’s  evidence that anesthetics may act by disabling   microtubules—thereby “disabling consciousness”.

So it’s not so crazy that Hameroff wondered   whether these molecules might actually have  some fundamental role in consciousness.

And then Hameroff read the Emperor’s New  Mind and reached out to Penrose with the   idea that microtubules could be a candidate  for this quantum information processor inside   neurons.

That was in the early nineties.

Since then Penrose and Hameroff and others   have developed some fairly involved scenarios  for how this might work.

The basic idea is that   information gets stored across one or many  microtubules, perhaps even across multiple   neurons.

The information in the form of quantum  bits—qubits—could be stored in a variety of ways,   for example the polarization  direction of individual tubulins.

So you have these quantum states that  are networks of entangled qubits,   and they are in superposition.

That means many  possible configurations of the 1s and 0s in the   qubits exist simultaneously across microtubules.

And then something happens to collapse this   superposition—some type of measurement causes a  single state to be chosen from the many.

And it’s   that moment of collapse that Penrose thinks is a  sort of proto-conscious moment.

So our conscious   experience would then be the sum of these  moments happening all the time across the brain.

By the way, Penrose also has a mechanism  for this wavefunction collapse.

It’s called   objective reduction, and it happens when  the superposition of different spacetime   curvatures corresponding to the different quantum  states reaches a certain threshold, causing one   quantum state to be chosen.

We did an episode  on objective reduction if you want the details.

The overall theory is called orchestrated  objective reduction, and the idea is that   the brain uses this wavefunction collapse as  part of its information processing mechanism,   and this leads to conscious experience and also  the ability to transcend classical computation.

Orchestrated objective reduction has  been around now for over 30 years,   and frankly very few people took it seriously—and  I’d guess most still don’t.

Some don’t agree with   the Penrose-Lucas argument, and fewer  believe that coherent quantum states   can persist for long enough inside the  messy environment of the cell interior.

Which brings us to the reason we’re doing  this episode now.

Evidence is emerging that   microtubules may exhibit interesting  quantum behaviors after all.

In 2013   evidence was found that microtubules  display large-scale quantum resonance,   giving them unusual long-range  electrical conduction properties   and potentially allowing them to work as memory  switching elements.

And just a few months ago,   a paper came out adding weight to this idea.

It’s  by Nathan Babcock and Phillip Kurian from Howard   University and their collaborators, and is titled  “Ultraviolet Superradiance from Mega-Networks of   Tryptophan in Biological Architectures”.

Now those meganetworks of tryptophan are   basically microtubules.

But to grok the rest of  that title we need to understand superradiance.

So, normally, if you radiate a bunch of atoms  or molecules with a bunch of photons, each might   absorb a photon, bumping an electron to a higher  energy level, and then radiate another photon as   the electron loses that energy.

When the emitted  photon has lower energy then the process is called   fluorescence.

The de-excitation of many excited  electrons takes a bit of time as they decay   randomly one by one, so fluorescent materials  glow for some time after being illuminated.

Superradiance is a bit different.

It can occur  when a group of atoms or molecules get into   excited states as an entangled group rather than  individually.

That can happen, for example, if the   spacing between excited particles is smaller  than the wavelength of the incoming photons,   then we can’t say which particle absorbed which  photon so we have to treat the collection of   particles as a single quantum superposition state  of all possibilities of the absorption matchups.

In any case, if the system of particles is in a  collective excited state then it tends to decay   collectively, and so photons are emitted together.

That results in a much brighter and shorter spike   of radiation than regular fluorescence.

Another  way to think of this is that the emission of   photons by one electron in the ensemble prompts  many identically excited electrons to decay,   amplifying the first electromagnetic wave.

Which, by the way, is also exactly how lasers   work.

Superradiance is sort of a scaled  down example of the laser phenomenon.

So, superradiance is a very quantum effect,  and these researchers claim to have observed   it in actual microtubules.

More precisely, they  claim to see superradiance in the tryptophan   amino acids within tubulin molecules  when radiated with ultraviolet light.

If this is right it would require  large-scale quantum entanglement   of tryptophan molecules across a microtubule.

Although the researchers of the paper didn’t  directly measure that entanglement or its source,   according to their models the amount of light  emitted was 100s of times that expected from   simple fluorescence, and so could have only  been possible through a superradiant process.

They also did a bunch of quantum  simulations to explore the effect,   and found that these entangled, excited states  can extend a long way along microtubules—far   longer than would normally be expected in  an environment like the brain.

This implies   that microtubules could at the very least  serve a role in cell signaling.

Finally,   they found that this coherence was more stable the  larger the microtubule and microtubule network,   which gives some hope that this effect might  actually appear in a real neuron.

On the other   hand, Penrose and Hameroff estimate that a good  fraction of all microtubules in all the neurons   in the brain need to be entangled to generate  human consciousness—and that level of warm,   wet, macroscopic entanglement  is pretty hard to believe.

So what are the implications of this?

Well forget  consciousness—even if microtubules contribute to   cognition, we may be way further from artificial  general intelligence than some believe.

Brains   have around 10^14 synapses.

AGI optimists have  said we’ll be able to simulate human intelligence   when we have computers capable of simulating that  many connections … which Moores law tells us will   be pretty soon.

However, if neurons are also  doing internal computation with microtubules,   and there are a billion microtubules per neuron  computing maybe a million times faster than the   neuron firing rate, then we may need to wait  quite a bit longer to make our first AGI.

And   at any rate, if orchestrated objective  reduction or anything like it is right,   no AI we make in the near future has any hope  of human-like consciousness.

If we want that,   and I’m not sure we want that, we’re going to  have to build our AIs on quantum computers.

After a good mix of science, philosophy, and  speculation, it's appropriate to finish with   a little disclaimer.

What the microtubules  superradiance really shows is that there are   potentially quantum processes happening in the  brain.

There is of course no claim in the recent   paper that this relates to consciousness at all.

Consciousness only comes into play if you accept   the arguments of Penrose, Lucas, and Hameroff.

We’ll leave it up to you to take a philosophical   stance there.

But I will say that I’ve always  been pretty dismissive of Penrose’s idea, but   this has me looking at it a bit more closely—and  the same is surely true of others.

But a lot of   work needs to be done to establish that  actual quantum information processing is   happening in microtubules and to even establish  a concrete mechanism by which this phenomenon   could generate our conscious experience.

On the  other hand, don’t you kind of feel like you’re a   massively entangled quantum state collapsing  under the buildup of superposed spacetime.