Talk:Integrated information theory

Unethical

Some of the text is copied verbatim from published papers. E.g. Google "We develop a detailed mathematical framework in which composition, information, integration" ... and this is not properly marked — Preceding unsigned comment added by 108.244.72.33 (talk) 01:43, 6 October 2014 (UTC)

Non-mathematical explanations / analogies needed

Fascinating theory. I'm not a mathematician. Help. Memills (talk) 03:49, 4 May 2012 (UTC)

Duplicate page

Page should be combined with duplicate page: Integrated Information Theory (IIT) — Preceding unsigned comment added by 144.183.224.2 (talk) 21:03, 8 August 2012‎

That's not a duplicate page – it's a (useless) redirect that was created automatically when this page was moved on 11 April 2012. It should probably be deleted (and I may propose that). —BarrelProof (talk) 18:04, 5 May 2013 (UTC)

More than a year later, and the article has not really improved and the topic seems to have insufficient evidence of notability. I have therefore proceeded to propose the article for deletion, saying "The topic does not seem sufficiently notable. It concerns an academic article that someone published that proposes a theory of consciousness that is not widely accepted or widely referenced in literature, has not received any award recognitions, etc. We should not add a Wikipedia article every time someone publishes an academic paper containing some idea that attracts a small number of other articles to discuss it." —BarrelProof (talk) 21:35, 30 July 2014 (UTC)

Removed delete article tag. Giulio Tononi is among the top researchers in consciousness studies. Other consciousness researchers suggest that integrated information theory is one of the best current theories of how consciousness is generated. For example, see Christof Koch's review. Notable. Memills (talk) 23:46, 30 July 2014 (UTC)

Very poor quality article

This is the kind of article that gives Wikipedia a bad name. It reads like the material the cold fusion people would write if Wikipedia would let them, or the way I would explain the ideas to a friend if my entire experience was a one-hour lecture about the subject. Please, somebody put the existing article out of its misery and write something like a real encyclopedia article. — Preceding unsigned comment added by 50.43.15.26 (talk) 02:16, 9 May 2014 (UTC)

Can blog posts count as a source?

Scott Aaronson seemed to pretty much refute this theory on his blog: http://www.scottaaronson.com/blog/?p=1799 128.97.68.15 (talk) 17:56, 28 May 2014 (UTC)

There is this paper with a critique: http://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1004286 --Arthurfragoso (talk) 22:31, 12 October 2015 (UTC)

Overall logic

There seems to be an incredible leap of faith equating information integration with consciousness. The article in no way develops this leap of faith. But, neither do the authors of the theory to any real level develop this. The logic is something like that: If humans do not integrate then they are not conscious. Therefore consciousness equates to integration. Oh well. My edits just somewhat toned down excessive claims and I will let others decide if this is a genuine contribution. — Preceding unsigned comment added by 165.124.149.74 (talk) 14:35, 27 June 2014 (UTC)

You are quite right. The originators do not properly address the "hard problem" in the theory of mind, just wave their hands and pretend the problem doesn't exist like so many uninformed materialists do. The proponents are notable and their attempt to develop a mathematical description of such complexities is laudable and, to my knowledge, novel. Let us hope they stick to what they are good at and backpedal on the philosophy they lack training in. — Cheers, Steelpillow (Talk) 18:55, 27 June 2014 (UTC)

They also don't seem to try to see if there's clearly non conscious systems that "integrate" "information". All sorts of chaotic systems have the property that localized differences affect the whole, and arbitrarily high phi can be attained in systems that nobody in their mind would call conscious (which Scott Aaronson demonstrated on a particularly mathematically convenient example which exhibits theoretically maximal integration). When called out on that, they started arguing that "integration of information" is a necessary but not sufficient condition, but a real number is not a condition. Faith leaping is the modus operandi; in the recent "3.0" paper they proposition that feedback is necessary and feedforward systems are never conscious. Which may seem intuitively reasonable, but you can implement an universal computer as a physically feedforward machine, and it makes no sense whatsoever if the consciousness would not be dependent on the actual code being run but would only exist if physical registers whose values are no longer necessary are overwritten by the intermediate values and be absent if all intermediaries are stored in place and never overwritten. Dmtrlk (talk) 20:04, 30 July 2014 (UTC)

IIT wiki page - Redesigned.

Hello all, I believe this theory is VERY promising. Unfortunately IIT has been victim to lack of exposure, which great lead to a great deal of miss-interpretation. Apparently Tononi & Koch are more interested in refining, improving and proving their theory than telling ppl about it. As they have been for the past decade.

This is where I step in: I have synthesized all of Tononi's work on IIT into one concise, approachable & accurate account of the Theory. All of this is in Tononi's words, up-to-date and accurate.

Please let me know of any format/syntax errors I might have missed. I will be adding the relevant images referenced in the page just as soon as I work out how to do that :P

I also hope ppl will point out to me problems with the flow of the article in terms of ordering - I have tried to make a streamlined account which begins from the 'ground-up' with thought experiments, axioms, postulates. Then turning to implications of the theory, empirical observations and what is left to be improved within the theory.

any advice is welcome. it is important that this page is approachable to the casual reader.

BTW - this is my first Wiki page :) feels good to do my part for this incredible theory :) — Preceding unsigned comment added by LozVieux (talk • contribs) 11:16, 27 August 2014 (UTC)

Welcome to Wikipedia! I agree, IIT is the best lead we have at the moment. The "hard problem" is a non-scientific red herring concocted to uphold the long-standing view that consciousness is supernatural. However, regarding the article; you've added far too much detail (ideally no statement should be made about the subject without a corresponding citation). I know that you think it's concise, but for a Wikipedia article it definitely is not. The article is just too long. Also, the style and tone are completely inappropriate; the article reads like something from a popular book, not an encyclopedia. You must not write in the first-person or specifically address the reader (ie using "you" or "us"). Your use bold and italics is also inappropriate. You should read WP:MOS. Also make sure to take a look at WP:FA to get a sense of what an appropriate article looks like. Also, although I agree that IIT is interesting, Wikipedia should not be used as a vehicle to increase the exposure of some particular subject matter to the lay public. Besides, generally speaking, it's not clear to me that it's useful or needed to increase the public exposure of scientific hypotheses; the validity and/or potential of this work will not be determined by the lay public. Wikipedia's only (proper) role is to provide as basic and concise an article as possible.Blacksun1942 (talk) 12:20, 27 August 2014 (UTC)

Hmmm....it appears that much of the content is lifted directly (plagiarized) from this source From the Phenomenology to the Mechanisms of Consciousness: Integrated Information Theory 3.0. Not good. Blacksun1942 (talk) 13:14, 27 August 2014 (UTC)

(edit conflict) I think this rewrite needs to be reverted. Primary sources should absolutely not be the basis of an article (and instead used only sparingly) and there should be much more secondary coverage of the ideas. Perhaps most troubling is that most of this edit appears to violate copyright as outright plagiarism of other people's work. I suggest nuking this article until it can come from as many (completely paraphrased) secondary sources as possible. czar ♔ 13:18, 27 August 2014 (UTC)

Proposed page rewrite - November 2015

Hi all. This is my first time contributing to Wikipedia, so I apologize in advance if I accidentally fail to follow established protocol/best practices. The page as it currently exists is seriously out of date. I have completed a draft rewrite, which you can find here. Is my user namespace the appropriate place for this, or should it be in a draft namespace? I would appreciate any comments on this new draft. I am hoping that some consensus arises and that I can replace the current article without stepping on too many toes. I have already talked to the article's initial author and have his blessing for this rewrite. So, thoughts? — Preceding unsigned comment added by Gbfindlay (talk • contribs) 04:55, 24 November 2015 (UTC)

I'm for it, because it seems to have more detail. But then I'm maybe not the best person to ask, I'm a mathematician, and have been following the page waiting for it to have sufficient detail that I can make sense of the theory. Your version has more content. Whether it actually reflects Integrated Information Theory, I'm not competent to judge. Make the change, and see what happens? DavidHobby (talk) —Preceding undated comment added 13:59, 24 November 2015 (UTC)

You bring up two good points: correctness and completeness. As for the rewrite's correctness, I am fairly confident of that. I work with the theory on a regular basis, so I hope that I haven't made any gross errors. Completeness is another story. In fact, I can assure you that my proposed article is NOT complete! Using only the information in the "Mathematics" section, you would probably not be able to calculate all the given quantities properly for every conceivable system. If you happen to be familiar with IIT, you will notice that little things like virtual elements and fixed background conditions to candidate sets are left out. But I am unsure whether Wikipedia is the place to roll out the theory in full detail. At some point I had to decide whether to make this an article about the math, which would be complete but probably too specialized/technical/overwhelming for a general audience, or to make it an encyclopedic overview of the essentials. I went with the latter, sacrificing a lot of completeness for accessibility. Unfortunately, IIT is hard to communicate in an elevator pitch. Beyond presenting technical challenges, it is also just requires a totally unintuitive way of thinking about certain things (moving from an extrinsic to intrinsic perspective of system behavior can be mind melting). And as with many other scientific/mathematical/philosophical theories, one probably cannot grok it from a encyclopedia article alone. At any rate, even though the proposed rewrite may not be perfect, it certainly should be more complete and more correct than the current article. I'm going to have a few colleagues who work with IIT give my draft a once over, and if they consider it an improvement I will go ahead and submit the changes sometime this week. Thanks for your thoughts! — Preceding unsigned comment added by Gbfindlay (talk • contribs) 18:28, 24 November 2015 (UTC)

Sorry, got bogged down with other stuff and finally put this revision up today. Better late than never? — Preceding unsigned comment added by Gbfindlay (talk • contribs) 17:18, 14 April 2016 (UTC)

The page rewrite

@Gbfindlay: overall this is a great improvement, thank you. I should like to offer two criticisms, hopefully constructive ones.

Firstly, there is perhaps more detail here than is required by an encyclopedia. It may prove better to summarise the formalisms in more condensed form, rather than present the raw equations. The reason for this is an encyclopedic one, in that there may prove to be little secondary literature discussing the individual formalisms.

My second criticism is more philosophical. The hard problem does not go away simply because one has produced a parallel formalism predicting the qualia of inner experience. The formalism indicates which quale ("qua-ley") is present but not what that quale feels like to the individual concerned. That, precisely, is the hard problem. It is frequently misunderstood by scientists, and the present state of the article (and presumably also Tonini) succumbs to that misunderstanding. If one is to make philosophical statements on Wikipedia then one needs a philosopher not a scientist to verify their correctness. I'd suggest that this aspect be rewritten along the lines of "Tonini claims that...."

— Cheers, Steelpillow (Talk) 18:36, 14 April 2016 (UTC)

@Steelpillow: Thanks for your thoughts. I would argue that the article makes no claim that IIT solves the hard problem as posed by Chalmers (Many people have co-opted the phrase to mean different things, I feel). Hence why the section is titled "Relationship to the Hard Problem", rather than "Solution to the Hard Problem." Rather, it's more like IIT acknowledges the problem and takes an approach which avoids the need to confront it. Maybe that could be made more clear.

As for the suggestion that there be less mathematical detail, I'm against it. The theory is by its nature mathematical, so a fair amount of detail should be included. DavidHobby (talk) 02:21, 15 April 2016 (UTC)

It is more like IIT acknowledges the problem and then avoids it from the outset by identifying the systemic complexity with the conscious experience. It is no more than a mathematical elaboration on the mind-brain identity theory (aka "Australian heresy") which so singularly failed to account for qualia. So many physical realists - and now information realists - assume that equating formal structures equates all their properties and resolves the Hard Problem. It does not. The Hard Problem is what is left over after you have done that. Science and applied mathematics deal with objective measurable properties. The qualitative properties of inner experience are subjective and un-measurable. This is the problem acknowledged in the article's opening remarks. Simply assuming that the two equate one-to-one, as IIT promptly does, is no proof of identity. The Hard Problem remains untouched. — Cheers, Steelpillow (Talk) 11:13, 28 July 2019 (UTC)

I agree that the article does not purport to address the "hard problem." That is a positive feature, because the hard problem is not even a scientific problem. The article points out however that IIT might be relevant data for philosophers discussing the hard problem. Personally, I think philosophers are barking up the wrong tree. It is simply a fact of life that I cannot directly experience your experience. Moreover scientists don't even try. Science always takes an objective or outsider's view. What they can try to do to model your subjective reports about your direct experience, which is exactly what IIT is attempting to do.

It is also relevant that purely philosophic outside criticisms of scientific theories are scientifically irrelevant. A scientific model or program be defeated in only two ways: 1. It stops producing testable hypotheses (or creates an endless stream of ad hoc splices), or 2. It is replaced by a better scientific theory.

IIT looks to me like a perfectly valid scientific research program that might (or might not) lead to real progress.

Burressd (talk) 17:43, 25 August 2021 (UTC)

PDF rendering fails

"Download as PDF" gives error: "Generation of the document file has failed. Status: Rendering process died with non zero code: 8" — Preceding unsigned comment added by 77.190.200.138 (talk) 15:55, 15 April 2017 (UTC)

Please do not top-post on Wikipedia, here we add any new section below the last one. Wikipedia's PDF rendering is buggy and only partly functional. I don't think it renders maths at all, or if it does, then the character set used in the maths is probably breaking it. Either way, it is not going to be fixed any time soon, see for example Help:Books. — Cheers, Steelpillow (Talk) 19:04, 15 April 2017 (UTC)

Spectrum? What "spectrum"?

The introductory section includes this sentence:

"The limitations on the physical system for consciousness to exist are unknown and consciousness may exist on spectrum implied by studies involving split brain patients and conscious patients with large amounts of brain matter missing." 2601:200:C000:1A0:6D0B:78E1:8950:3FB4 (talk) 21:49, 1 November 2022 (UTC)

The sentence is ungrammatical, because in English the word "spectrum" should be preceded by either the word "a" or the word "the".

But in any case it is entirely unclear what the word "spectrum" means here.

I hope someone knowledgeable about this subject can fix this. 2601:200:C000:1A0:6D0B:78E1:8950:3FB4 (talk) 21:52, 1 November 2022 (UTC)

Cause-effect space

The subsection on Cause-effect space states that the set of probabilities of the before and after states of a simple binary element is a point or "star" which represents a concept. Can somebody please explain, on what basis does a concept necessarily comprise before and after states? Tononi is notorious for arguing that a sufficiently large and complex computer memory may be conscious, even if it is static like an unplugged USB stick. There appears a deep inconsistency between the notions of stasis and causal before-after. This bears explanation. Or, if I am missing something, can somebody please explain what? — Cheers, Steelpillow (Talk) 19:21, 19 November 2022 (UTC)

Puffery and weasel-words in lead paragraph

For the second time, I have removed In principle, a mature and tested theory of IIT may be capable of providing a concrete inference about whether any physical system is conscious, to what degree, and what particular experience it is having. from the lead paragraph.
The sentence has too much in the way of puffery and weasel-words to convey any actual meaning. It doesn't say anything, it just hypes up what might be to make this look more important than it is. 217.180.228.138 (talk) 18:43, 21 September 2023 (UTC)