Can math explain consciousness?
ฝัง
- เผยแพร่เมื่อ 26 ก.ย. 2024
- This is an entry for a contest of Theories of Everything with Curt Jaimungal: / theoriesofeverything (specific video for the contest is • Physics & Consciousnes... )
by
Rodrigo Coin Curvo
&
Alexander Maier
This video is a remake of: • The Mathematics of Con...
Read more about Integrated Information Theory and the #neuroscience of #consciousness: www.scholarpedi....
Also, check out Rodrigo's entry to SoME1, which goes more into detail regarding the math behind IIT:
www.youtube.co....
--------------
Attributions:
Icons from flaticon.com: scissos, newborn and world-book-day (by Freepik); AI and bedtime (by photo3idea_studio); brain 1 and 2 (by Smashicons); deer (by tulpahn).
Images from Wikipedia: Giulio Tononi, Kristof Koch, Ernst Mach's Innenperspektive and Indian elephant.
Soundtrack by Alexander Maier
--------------
References:
pubmed.ncbi.nl...
pubmed.ncbi.nl...
pubmed.ncbi.nl...
pubmed.ncbi.nl...
Animations were created using Manim.
Manim is a free and open-source project originally written by Grant Sanderson ( / 3blue1brown .
Manim is now maintained by the Manim Community and released under the MIT license. www.manim.comm...
Thank you so much for your submission! (final video announcement with winners / runner ups out now, by the way)
What an honor to be chosen as one of the winners!
Thank you so much again for organizing this awesome, interesting and fun event. The idea of having a #SoME-type series of videos for consciousness science is genius. Looking forward to #PaCE2!
@@astonishinghypothesis Of course! Your video was great Thank you so much
A brief *philosophical* take on IIT:
Things exists only so far as they have causal power: either affecting stuff, or being affected by stuff themselves. A "thing" that can’t be affected by anything and isn’t affecting anything in its turn is not a thing to start with-it’s nothing! A rock, the sun, quarks, carrots, gravity, water, laws, etc they can and are affecting stuff, and in turn, they can be affected by stuff. Hence they exist, but they exist for other stuff! They exist as far as other things are concerned. So, from the start: for consciousness to exist it must meet this requirement of having some kind of causal power. But a rock, the sun, quarks, carrots, gravity, water, laws, etc aren’t considered conscious. And if these things aren’t conscious, then for brains to be conscious they must require something extra. Consciousness must posses some property besides mere causal power. IIT says that this “something extra” manifested by conscious systems (like brains) is: causal power upon itself . When something has causal power upon itself , being able to constrain it’s future states (and possible past states in which it could’ve been), then it exists for itself! "Existing for itself" in the eyes of IIT is the same thing with "being conscious". The brain has this property in virtue of its phenomenal degree of integration through feedback loops-but it’s only present at a very particular spatio-temporal grain (NOT all the brain and at all time frames). IIT has a mathematical framework that measures the degree of integration (which is the same thing with the degree of "existing for itself" or "consciousness") in various systems and it ascribes them a number called "phi". A system having 6.78 phi is vastly less conscious than a system having a phi of 368. A system having a phi = 0 is unconscious. IIT says that computers with a feed forward architecture of information processing have a phi = 0; meaning that they’re not conscious, irrespective how intelligent they are. But if you make a silicon computer that has a phi greater than 0 then that computer will have a feeling. What kind of feeling? Again, IIT in its current form doesn’t say but like every other scientific theory out there it assumes that inner states could theoretically be mapped onto an abstract mathematical space so that if we ever develop sufficiently powerful technology we could then learn the correlations between inner experiences and their respective mathematical construct. So, if we could "scan" brains to the relevant degrees of accuracy we could use an AI to approximate what those brains are experiencing.
In the past the proponents of IIT we’re willing to accept that the underlying “stuff” of nature might be conscious in some primitive way-particles (quantum field excitations) might actually be very rudimentary “glimpses”/”sparks of qualia”. In which case IIT takes a weak form, in which it only describes how these primordial “sparks” can come together to form a higher order entity (otherwise, by simply putting them together you’d only form “piles” or “ensembles”, like chairs, buildings and planets, not entities like our conscious mind). I don’t know if they’re still willing to accept weak IIT, but i’m liking both the weak and strong versions.
Dude, thank you. This is exactly what I was looking for. I hope I can find more philosophical interpretations like this.
There's the idea that consciousness might lack any casual power at all, there are reasons to think that's the case. In fact, panpsyquist theories are the ones that make easier it to explain why consciousness exists with no casual power, it exists because it just does. No need to explain why evolution produces consciousness if consciousness just emerges from information integration (or whatever) per se.
The only thing is that I find incredibly weird the idea that there's no mental causation since we are trying to explain why we have consciousness. If consciousness causes nothing of our behavior, it seems that we just started talking about our subjective experiences by sheer luck, which is absurd.
The only alternative that I can come up with is that we are actually deeply wrong about our psychology.
Tangentially, should we imagine P-zombies stopping to wonder about conscious experience? It seems absurd, but if we don't think they would, then we are already recognising that consciousness has causal power, even though we haven't observed unambiguous examples of mental causation. So we should not just assume it.
It doesn't really matter, P-zombies are just a mental experiment, I don't know if these ramblings just come from taking the mental experiment too seriously.
I'm into the idea that consciousness is the only thing fundamental, that spacetime comes forth out of consciousness. Your consciousness is a fractal representation of the greater consciousness. It's just there "observing", the rest is the brain doing some processing and is probably really well suited to be a portal to the greater consciousness. But a fish, ant, virus or rock are also "conscious" it's just so far removed from our conscious experience, that we say it's not the same.
We have this interface and we see the world through it. But it's just an interface and doesn't tell much about what is going on in reality. Just like this text that you're reading has meaning and feels real but if I show you just the circuit board of your phone right now, that is reality. You wouldn't be able to tell that this text is the "fake" and the charges going through the circuit is "reality" since the interface gives way more meaning to us than just electrons moving around between some silicon.
It's kinda like panpsychism or idealism but the thinker of the theory (which my explanation probably butchered) Donald Hoffman calls it conscious realism. What I find most interesting is not as much as the theory itself and it's implications, but the method of how this theory came into existence. Basically using game theory on conscious agents and see if evolution would discard raw reality and make an interface
A unique formulation of *metamathematics* can “explain” consciousness as a self-referential mechanism.
Hello, I hope you did see the CTMU explanation PaCE1 video.
Seems like you follow CTMU closely as well right?
@@danielvarga_p Yes, I do!
You can always define a self-referential operator, or self-referential systems, but how do you connect the math with reality? There has to be some kind of empirical framework that posits theoretical claims that connect the math with reality.
@@didack1419 Excellent point. The "math" - when defined and developed fully self-referentially - is equipped with sufficient generality to contain "physical reality" (i.e., the physical, chemical, biological, etc. examinations of reality) as its "display domain". Coupled with this domain is a "processing domain" in which the mathematical structure of reality resides.
@@Self-Duality Bollocks mate. "The math" can never be self-referential (Russell's paradox). The phenomena maybe. But then the noumena cannot be "math" and this whole insane idea falls apart. It is semantics otherwise. If you leave of subjective qualia from _your_ definition of "consciousness" (which is just behavioural) you have not explained one iota of my definition of consciousness, which I claim exists, at least in me. Not sure about you.
Excellent
Awesome!!!!!
@14:57 did not end up answering your title. Using IIT in the way you suggest (or in _any_ way) can only ever describe behaviour. But you noted that at the very beginning around @2:00 min in, so.... Math cannot do it would be a valid conclusion. Math can use _past_ knowledge (in an already conscious person) to infer they may or may not be conscious now. Nothing about qualia. Nothing about strings of toilet rolls or ant colonies that would 'pass' as conscious by your math. At the end of the day it is curve fitting. Curve fitting (statistical modelling) is only valid for inferring form the past. So can never be telling you about fundamental ontology. See the work of Yasaman Razeghi and Prof. Sameer Singh --- the ANN are all about memorizing data sets, nothing else. Zombies. Which can only ever be used for statistical inference. But statistics is never accounting for ontology. (This is why quantum field theory is not ontology, it is non-Kolmogorv statistics - non-distributive lattices.)
Billiard balls that interact with each other are not falling in love or experiencing the colour red. IIT is so silly.
If Set B is under Set A, is it possible to produce Set A from B? Is my reasoning right that mathematics is just a subset of our consciousness? So can we think of this problem as a subset being able to produce the original set in which it belongs?
I don't think math is a subset of consciousness; math is basically just logic, and logic seems to be something fundamental to the universe since as far as we know the universe is completely logical and can be described with math. Since consciousness is part of the universe I am assuming it can be described with math.
@@Bob13454 in the context of your reply what does "logic" or "logical" mean exactly? Does it mean reasonable, rational, or the argumentation logic?
@@lorenzomizushal3980 It means if a=b and b=c then a=c
Let me deep dive this.
explanations is not the same as the territory so of course you can explain it through math just as philosophers explain through language. but symbols will never be to totally explain consciousness because it is consciousness. mathematics IS consciousness. but it’s fun to attempt. you may be on to something.
The idea of explaining real things through math is that the math represents the real structure of those natural things. It's possible that math can indeed represent the real structure of consciousness.
Confusing the model for the phenomenon is called reification fallacy or map-territory equivocation. I don't know how math can literally be consciousness, given that math is just an extrinsic process of modelling done by organisms (us) that have indirect access to reality and the intrinsic phenomena that occur in it.
You start off ok, defining the difference between objective behaviour and subjectivity. But then you balls it all up by going to IIT, which is all objective, nothing subjective. All you are doing is using _your_ subjectivity and then taking the IIT data+ mappings to infer someone else had similar qualia-filled experiences. How do you know though? Because you figure your qualia are as real as theirs. So that is not what a scientist calls an "explanation". Science can _never_ refer to a subject, except as a 'test vehicle' or whatever. If you cannot eliminate a thinker form the situation, you are not doing objective science, and have zero scientific account for inner subjective qualia-filled consciousness.
Another way to put it (though you aught to read the literature) is that no one can use the IIT + brain scan data to learn anything about a patient's qualia. It is all inferred from your oen qualia, whcih no one else knows about. You see the problem?
The valid inferences are "that some qualia was likely experienced". Which is ok for science. The PROBLEM is _you_ have to have a conscious mind to make the inferences to the qualia you are claiming the patient was experiencing. Do you even get what I'm writing about? If not, you could be a zombie. IIT is fine for detecting the objective manifestations of consciousness. But tells you nothing about consciousness. (Metaphor: the rays of light from a lamp are not the lamp. By measuring the rays I still have no idea what it is like to be the lamp. Was it all bounced off a mirror, etc.)
All science starts with your own subjectivity. Unconscious matter cannot do science.
Your objection seems to be (the common notion) that you cannot make any inferences from your own subjectivity to that of others': If we both look at the same color and I see "red", how do I know whether you see the same "red"?
Check out Nao Tsuchiya's work on the Yoneda lemma for the answer. Mathematically, our experience has to be the same, if we agree on how the redness we observe relates to all other colors (that is, if we agree on color space).
Mathematics doesn’t explain anything - it’s not a science
Math explains stuff in the sense in which physical theories explain stuff using math. You can use math to describe reality and then to predict things, that is what makes a scientific theory.
ITT doesn't have strong empirical evidence (none as far as I know), but the idea is still scientific insofar as it is a hypothesis about reality.
It's just not yet testable, (presumably) not unlike other physical theories though. Or maybe it is completely untestable if consciousness does nothing.
I've seen that Chalmers also thinks that consciousness might cause the collapse of the wave function, if that's true, then that might make it testable.
@@didack1419 Petra Kann is sorta' right. Mathematics _describes_ aspects of reality that are "mathematical" (not a tautology, just a shorthand for {logical, numeric, geometric, algebraic, etc.} ) This is important. The description, the epistemology, is not the ontology. If you think it _is_ then you are some sort of Idealist, so not doing science à la Aristotle to Kant to Einstein.
@@Achrononmaster But that's literally what I said, mathematics explains things because physics requires math to do the explaining, not because the math can do the explaining by itself. Math (the process of highly-formalised abstract modelling of things we perceive or conceive) produces descriptions of reality, which we experiment with, and that's literally physics.
I don't see how I'm confusing the explanandum for the explanans. In fact, it almost seems to me that you are the one confusing the two because of your seemingly implicit admission that "ontology is reality", while any ontology is just a description of reality, more or less accurate.
Fuck you mean it's not a science
The the brain itself would then be 1 and 0 simultaneously as neurons inside the brain would be 1 or 0, which makes the whole.