That's a fair assessment, and in line with Watkins' thinking.

What I've seen of Friston's publications, including videos, shows a certain humility, or at least a reserved surprise at the general applicability of the mathematics. Not at all unseemly.

It's fascinating to watch, a real-time demonstration of the methodology of research programs. Overreach would be expected as the new model settles.

So if I get it straight, your main objection is the way Banno presented Friston’s idea here on this thread? Using the "dark room problem"? You find it misleading? — dimosthenis9

Banno compounded the confusion for sure. But my comment was directed at the dark-room problem itself. It is motivated by a misunderstanding of the theory it intends to question. And the misunderstanding is so obvious that I wonder how it got started.

Maybe someone else can spot it, but I see no reference in the paper to the supposed source of the dark-room problem. The authors just say...

A recurrent puzzle raised by critics of these models is that biological systems do not seem to avoid surprises. We do not simply seek a dark, unchanging chamber, and stay there. This is the “Dark-Room Problem.”

... so the suspicion I voiced was that Andy Clark was pulling a smart career move to get some kind of controversy going. The dark room problem was an invention to draw attention, create a dialog, even if the lobbing of this "recurrent puzzle" into the conversation was going to be a patsy toss easily batted to the boundary by the home team.

Then as far as Banno goes, it seemed curious that he should pluck this paper out of the past just when I happened to be promoting Friston as having arrived at neuroscience's theory of everything in another thread. He will say that is an innocent coincidence of course.

And it is also curious that instead of pointing to the obvious misunderstanding of the free-energy principle used to motivate the dark room problem – which I feel he gets – he said his real interest was...

But why is minimising surprise the very same as living longest? — Banno

...which is the promotion of an even lamer quibble. How could one pretend to be baffled by that? And why did Banno ignore my point that the biologically correct reply would be that living long ain't so much the point as reproducing the most successfully. (Try maximising that by retreating into a darkened room with only yourself and your sexual fantasies.)

So the paper itself is a bit of a time waster. And Banno played his usual game of stirring the pot, claiming interest without being interested enough to spend 10 seconds clicking on a link and scrolling to a time stamp.

I'm not complaining of course. It is really funny that Banno's essential complaint against theories of everything are in fact exactly what one finds once one delves into the actual maths of the free-energy principle. Poetic justice when you blow yourself up with the very mine you tried to lay.

So in your opinion, and in a few lines what exactly this theory tell us about consciousness? Or what it implies at least as to rephrase it. Why you find it so huge? — dimosthenis9

As I said, it puts neuroscience on its own secure mathematical basis. It gets neuroscience off computation - the Universal Turing Machine formalism - as the general theory of everything it has been employing. It underwrites the whole shift back to an embodied, enactive and semiotic approach to mind science.

I was asked to write an opinion piece for the 20th anniversary of a major neuroscience journal on what I thought was the biggest thing of the past 20 years. My answer is exactly this. The construction of a mathematical formalism with the generality of any other "physical mechanics" that mind science can now call its own.

Well, not quite. We want a theory that rules out things that are contradicted by the evidence. — Banno

I am sympathetic to your attitude towards totalizing theories. But there is a difference between a general unifying idea and a detailed treatment of a subject. — SophistiCat

Yep. A totalising theory must still have some counterfactual impact on our thought. A new paradigm must be able to knock down the one it means to replace.

So Darwinian evolution killed a creationist account of life on Earth, for all practical purposes. And it would be nice if Friston's Bayesian Mechanics kills off Universal Turing Machines as the general paradigm framing mind science.

. It gets neuroscience off computation - the Universal Turing Machine formalism - as the general theory of everything it has been employing. It underwrites the whole shift back to an embodied, enactive and semiotic approach to mind science. — apokrisis

Yeah I read that in your previous post also.
So in general, if I get it right. For example in the "mind/ body problem" (physical /no psychical) you would use that theory as an argument in favor of the idealistic no physical nature of the mind. Right?

Not that it demonstrates or proves that, but I hope you get my point.

it seemed curious that ( Banno ) should pluck this paper out of the past just when I happened to be promoting Friston — apokrisis

No surprises there. The Dark Room Problem was mentioned by Friston in the questions of a video you posted. The article in the OP came up on a search, and as the length of this thread shows, was worthy of its own discussion. You can have the credit, if it is important to you.

So far as the origin of the problem goes, when I went looking earlier, the earliest mention I found was the article in the OP - other articles refer to it. Others might have more success. Clark may be to blame, and if it was a "smart career move" that is because it spiked interest in the topic, so good for him.

And finally, as for "why is minimising surprise the very same as living longest?", this is an rough slogan fo the sort mentioned by . And in this case the details remain to be settled.

So in general, if I get it right. For example in the "mind/ body problem" (physical /no psychical) you would use that theory as an argument in favor of the idealistic no physical nature of the mind. Right? — dimosthenis9

I'm not sure I understood your question. But I think the answer is that the computational approach to mind is inherently Cartesian, hence dualist and representationalist. Friston's approach is inherently semiotic, hence triadic and enactive.

And you can claim that semiotics sounds sort of idealist. But then it is an idealism that is a pragmatic realism. So not really idealism, but a way of fixing the fundamental issues around knowing the world that Kant raised.

Idealism is a broad church. But if you mean idealism in terms of some synonym of soul, spirit or consciousness as a monistic substance - one that stands opposed to matter as the other candidate monistic substance underlying reality - then no way am I making any argument in that direction.

What I am saying is that a triadic systems perspective - like Peircean semiotics, like Friston's Bayesian mechanics – does the best job of dissolving the tensions of the Cartesian notion that is "the Hard Problem".

Dualism says there are two distinct realms of being and no good causal explanation of how they might then interact.

Triadic logic says there is an interaction. And then two distinct realms are what develop out of this fundamental connection.

So an ontology of opposed substances - matter and mind - is replaced by an ontology of a self-organising dichotomy. The unity of opposites which is the brain modelling its world in terms of its self-interest.

The self establishes itself in the world by becoming causally opposed to this world. An organism gains autonomy by finding a way to impose its desires on an environment.

This trick is achieved by the interaction that is the modelling relation. Friston puts that into the familiar language of differential equations.

Peirce described it in logic. Rosen described it in category theory. Friston uses now the kind of maths that even engineers can understand.

Think about it. UTMs are finite state automata. Almost immediately engineers could see how to turn an algebra of information - a digital encoding - into real world programmable machines.

But then computer science was also tinkering away with these other things called neural networks. They had some rules of thumb, but no fundamental theory. It was a promising project but stumbled along in the background.

Now Friston offers a geometric approach to information. It is all about trajectories on attractors. Instability, self-reference and complexity are all built in. And yet still, the chaos can be tamed enough to model it with a set of differential equations – providing you can apply the right constraints, the right learning environments.

So people can complain that they don't see a theory of consciousness in Friston's Bayesian mechanics. They can complain that the surprise a human feels is nothing like the surprisal - the free energy metric – that Friston's formalism minimises.

Fine. They just show that they don't even realise how deeply they are stuck in a Cartesian framing of the whole issue. They don't even begin to get the paradigm change being talked about.

No surprises there. The Dark Room Problem was mentioned by Friston in the questions of a video you posted. — Banno

You read the comments on YouTube videos? And yet you say you are too old to watch the videos themselves?

Meh. A curious approach to scholarship indeed.

And in this case the details remain to be settled. — Banno

How so exactly? Be specific and not general.

You read the comments on YouTube videos? And yet you say you are too old to watch the videos themselves? — apokrisis

I don't watch videos - much. It seems inevitable that the young people take this easy rout, but I'd much rather a paper; doubtless it's habit, but the arguments are easier for me to follow, they have citations, and I can flick back and forth between paragraphs without fiddling if I miss something.

The video you shared caught my curiosity, I flicked through it looking for a paper, and found mention of the dark room.

Look, Apo, I'm flattered and all that you feel the need to make me the topic of this thread, but I don't owe you explanations or replies, and I'm sure others are finding this tedious.

Idealism is a broad church. But if you mean idealism in terms of some synonym of soul, spirit or consciousness as a monistic substance - one that stands opposed to matter as the other candidate monistic substance underlying reality - then no way am I making any argument in that direction. — apokrisis

Well idealism is a broad church indeed. If I could be more specific I meant mostly about the issue about the material or not nature of the mind.

I stand with the non material nature of the mind, which for sure as to exist requires material (brain) but on its own it's not something material also. Like material creates something non material. Interacting together. That's why I was so curious about.

Triadic logic says there is an interaction. And then two distinct realms are what develop out of this fundamental connection. — apokrisis

Interesting. So it makes a distinction indeed and supports the interaction.

They can complain that the surprise a human feels is nothing like the surprisal - the free energy metric – that Friston's formalism minimises. — apokrisis

Well I also complain about that as to be honest. Wasn't very convinced but as I wrote it does have interesting things.And well your response was really useful.

but I'd much rather a paper — Banno

Sure. And that takes about 10 seconds on Google. So....

Acting on the environment by minimising free-energy through action enforces a sampling of sensory data that is consistent with the current representation. This can be seen with a second
rearrangement of the free-energy as a mixture of accuracy and complexity. Crucially, action can only affect accuracy. This means the brain will reconfigure its sensory epithelia to sample inputs that are predicted by its representations; in other words, to minimise prediction error.

fil.ion.ucl.ac.uk/~karl/The%20free-energy%20principle%20-%20a%20rough%20guide%20to%20the%20brain.pdf

The equation you seek to avoid is next to that caption in Box 1.

Look, Apo, I'm flattered and all that you feel the need to make me the topic of this thread, but I don't owe you explanations or replies, and I'm sure others are finding this tedious. — Banno

The familiar gambit. But look Banno, my complaint that you demonstrate disingenuity by claiming to be interested in something and yet failing to respond on that something as soon as it makes you look flat-footed. That is what I'm sure me and others find tedious.

Even on the simplest question, you play the wombat defence. You say a matter of interest is "why is minimising surprise the very same as living longest?".

And indeed that is not the argument a biologist would make, although an information theorist might.

So for the benefit of your interest, I explained. Darwinian selection would not be thought to maximise life span but maximise reproductive fitness. (Which is another one of those pesky risk-reward, complexity-accuracy, trade offs that the equation you don't want to look at models.)

Seen in that light - the need for an organism to get busy with the jiggy - the silliness of the dark-room problem because blindingly apparent. What the free-energy principle minimises is the prediction error in following the life course that achieves that goal. The Bayesian Brain becomes the Lagrangian of optimised reproduction.

Now you might not like the fact that you didn't immediately see this obvious answer. But getting in a huff, going off to fume while you inspect your rotting garlic and spider infested terracotta, is entirely you making the discussion all about you and your ego.

Blame me and claim it to be a form of flattery. But I know that you know that you know better.

All you have to do is respond to informed replies with a little more scholarly integrity. It's not as if I am setting a high bar even there.

The familiar gambit. — apokrisis

Well, you keep falling into it. It works - utilitarianism in action; you should be pleased.

I'm not trying to present a comprehensive theory here - I just pointed to an interesting article. I've nothing to defend, and no more to say unless someone says something interesting...

Which hasn't happened this page.

I stand with the non material nature of the mind, which for sure as to exist requires material (brain) but on its own it's not something material also. — dimosthenis9

That leads to the symbol grounding problem - how does a code do its job?

A code is "physic-less" not because it has no material base, but because that material base is constrained to be a single standard cost. So a symbol is physical dimensionality constrained to where it becomes a dimensionless point. Just a bare mark. Like a letter on a page, a spike on a wire, the switching of a logic gate.

It takes as much, and as little, effort to utter any grammatically-structure sentence. Each word is just a puff of air, whether I speak the ravings of a lunatic or recite a play by Shakespeare.

So codes have to have some physical base so they materially exist. But that entropic cost falls out of the equation if it is some tiny and inconsequential sum that is constant, no matter how random or magnificent might be the meanings embodied by symbol sequence itself.

In that way, the brain-mind does create a "safe place" which can stand outside the physical world so as to regulate the goings-on of that physical world.

For all practical purposes, a code puts the organism outside of the second law ruled world. For all practical purposes, the brain implements an idealist ontology - a realm that is like a virtual machine with absolute freedom of action.

But - as Friston models, and the second law demands - there is an electricity bill to pay to keep the lights on. Even small costs can add up. The brain-mind's idealism only makes metaphysical sense if the benefits of the free imagination are coupled to the costs of that free imagination over the long run.

Any model of mind has to include the two sides to the equation. Which is where we get back to the triadic systems view of consciousness.

Well, you keep falling into it. It works - utilitarianism in action; you should be pleased. — Banno

Ah, you say, it tis but a scratch. More video I'm afraid. But you've seen this one.

Which hasn't happened this page. — Banno

Long life vs reproductive fitness? Hello. Anyone there? I can hear your heavy breathing on the line. Come on out and defend. Don't just play for stalemate. The invisible audience that you think is hanging on your every post is waiting to see something more interesting.

:razz: I'll bleed on you!

Thanks for that link: The free-energy principle: a rough guide to the brain? I notice that it was published in the "Opinion" section.

If you are satisfied that this stuff also explains evolution and consciousness, then I'm happy for you. For reasons already explained, I'm not so keen, but open to the possibility.

Note that it's not "long life vs reproductive fitness" with which I take issue, but "why is minimising surprise the very same as living longest?" Thanks to you and others, I've a better understanding of what's involved. I'm happy to await further research.

I don't think I'm the only one here who thinks that thinking this already settled is overreach.

Are we playing "posts-last-wins" now? I'm good at that.

If you are satisfied that this stuff also explains evolution and consciousness, then I'm happy for you. — Banno

Did I say explain or ground explanations?

Note that it's not "long life vs reproductive fitness" with which I take issue, but "why is minimising surprise the very same as living longest?" — Banno

So all you had to do was tell me you agreed. Instead you took the disagreeable path of pretending not to have heard.

I don't think I'm the only one here who thinks that thinking this already settled is overreach. — Banno

Again, if a field finds its grounding, then things are not settled but properly (re)started. Don't misrepresent what I say so as to be able to claim to be right about what I never said.

And stop appealing to your invisible audience. You sound like a politician. Its Trump's go-to gambit for a reason.

Are we playing "posts-last-wins" now? I'm good at that. — Banno

So you say. I'll be the judge of that.

So all you had to do was tell me you agreed. — apokrisis

Apo, I did, several times.

I just made a cheese toasty, using a flatbread that was past it's "best by..." date, but which had been in the fridge, and had no obvious signs of mould. I put on the new Genesis album, quite loud, unaware that it was a rehash of their old stuff. But I enjoyed it, despite it lacking surprises.

There's the rub' it's just not obvious how what we do is result of our avoiding surprises. But it might be.

Apo, I did, several times. — Banno

:lol:

I just made a cheese toasty, using a flatbread that was past it's "best by..." date, but which had been in the fridge, and had no obvious signs of mould. I put on the new Genesis album, quite loud, unaware that it was a rehash of their old stuff. But I enjoyed it, despite it lacking surprises. — Banno

Now you just want to make me sad. Toasties and Genesis FFS's?

There's the rub' it's just not obvious how what we do is result of our avoiding surprises. But it might be. — Banno

Not obvious to whom exactly? The Peircean collective wisdom?

Well why does Friston command an h-index of 253? (h-index of 20 = good, 40 i= outstanding, 60 = truly exceptional; h-index of typical PF poster, some comical negative value.)

You keep making these third person claims about your first person views. Again, that's political rhetoric not supported argument.

So show me why my choice of Genesis drops out of Friston's equations.

why does Friston command an h-index of 253? — apokrisis

Because it's interesting, of course; but perhaps someone who thought in terms of the Peircean collective wisdom would mistake a high h-index for truth, or an approximation thereto.

So show me why my choice of Genesis drops out of Friston's equations. — Banno

Stan Salthe's lifecycle model of ecology will do that. It is the kind of more detailed theory that you demand.

You say it yourself. You've got old. Brittle and senescent, to use the technical terms. That you would have Genesis on the turntable, rather than Black Midi or Connan Mockasin, speaks to your reduced capacity to deal with environmental novelty (even if you have the other side of the trade-off in the conviction of your certainties, the wisdom of a lifetime of evermore entrenched habit.)

Because it's interesting, of course — Banno

Yeah. Nobel prizes go to the most interesting theory of the year. I'd forgotten that's how it works.

but perhaps someone who thought in terms of the Peircean collective wisdom — Banno

But that is the pragmatic theory of truth you support yourself. The limit of exhaustive rational inquiry.

It is why you keep making plaintive appeals to an invisible audience who could only agree with your own good commonsense.

"No one agrees with you, Apo. Everyone agrees with me! Banno!!!"

(Tip: To be read in a Trump accent.)

So I've watched the video and had a bit of a look around and now this thread seems to be mostly people arguing past each other (including my first post to be honest) and arriving at the fairly obvious conclusion:

A theory of how consciousness works would of course be very interesting, but I think we can agree that it would require more details. Minimising free energy is a very high-level idea, an abstraction at the level of thermodynamics and definitions of what constitutes information. I think we can agree that consciousness is more complicated than heat. Still, it's useful to have abstract frameworks because they guide the more detailed hypotheses, and limit the possibilities that are worth considering.

Then again, if he got something essentially right, then these kinds of big-picture narratives can be valuable as setting directions for future research and providing an insight into large-scale patterns. — SophistiCat

:up:

Minimising free energy is a very high-level idea, an abstraction at the level of thermodynamics and definitions of what constitutes information. I think we can agree that consciousness is more complicated than heat. — the affirmation of strife

It just tells us what kind of physics to root our life and mind science in.

Science believes it is constructing a hierarchy of theories. Neuroscience is explained by biology, biology by chemistry, chemistry by particle physics, particle physics by quantum mechanics.

The paradigm shift is to instead see that biology is rooted in the physics of thermodynamics. In particular, the thermodynamics of “far from equilbrium” dissipative structures, So not the science of material atoms, but the science of evolved or self organising physical complexity.

And even particle physics and cosmology themselves have made this paradigm shift. So even fundamental physics has abandoned simple atomism for the more complex view that reality as a whole is a dissipative structure,

Hence Friston is just doing the work to catch neuroscience up.