This was an email thread initiated by Warren Mansell. It is preserved here for the record and as an artifact of communication amenable to PCT analysis.
Participants:
Warren Mansell warren.mansell@curtin.edu.au
“Madhur Manglam” mmangalam@unomaha.edu
Richard Kennaway richard@kennaway.org.uk,
Richard Marken rsmarken@gmail.com,
Henry Yin hy43@duke.edu,
Vyv Huddy v.huddy@sheffield.ac.uk, [removed after 2025/10/23]
Tauseef Gulrez tauseef.gulrez@curtin.edu.au,
RupertYoung rupert@perceptualrobots.com,
Ty Roachford ty@tyfoodsforthought.com,
Brian Fleming brianthebarkeologist@gmail.com,
Ehud Ahissar ehud.ahissar@weizmann.ac.il,
Bruce Nevin bnhpct@gmail.com,
Peter Osei posei2020@fau.edu,
Tauseef Gulrez gtauseef@ieee.org,
Ty Roachford troachford2013@fau.edu,
Kent McClelland mcclelgrinnell@gmail.com,
Philip Farrell PHILIP.FARRELL@forces.gc.ca [After 2025/10/20 replaced with farrell.pse@gmail.com]
Henry Yin hy43@duke.edu [after 2025/10/20]
Warren Mansell warren.mansell@curtin.edu.au
Oct 18, 2025, 11:37 PM
Dear colleagues,
I’ve recently been in contact with Madhur, who is a researcher in biomechanics and has a longstanding focus on challenging the same issues of group statistics, probability and prediction that we have been challenging…
Talk to you soon,
Warren
Warren Mansell warren.mansell@curtin.edu.au
Sun, Oct 19, 2025, 2:28 PM
Some of you will like these too…
https://www.researchgate.net/publication/391807467
https://www.researchgate.net/publication/390546288
Part of me is nervous that if Madhur turns his sites onto PCT, we are doomed, and part of me is quietly confident that we are safe, given Bill’s and Rick’s spadework!
Richard Kennaway richard@kennaway.org.uk
Oct 19, 2025, 3:40 PM
Wow!
– Richard
Madhur Mangalam mmangalam@unomaha.edu
Oct 19, 2025, 6:55 PM
I’ll take that as a complement. Glad to have these mutual feelings about Fristothermosycophantogenesis.
Warren Mansell warren.mansell@curtin.edu.au
Oct 19, 2025, 8:40 PM
Hi Madhur,
Agreed, I’m sure you’ll get more too…
Richard Marken rsmarken@gmail.com
Oct 19, 2025, 11:15 PM
Madhur’s “Idiot’s Guide” and “Exit Manual” are great fun; they’re funny because they’re true. But one line from the Abstract of the “Emperor’s New Pseudo-Theory” paper caught my eye because it was similar to something I said in the conclusion to my IAPCT talk. The line in Madhur’s paper is:
Despite its institutional prestige, the FEP has failed to produce a single novel, testable, and non-retrofit prediction about brain or behavior—while simultaneously crowding out constraint-based, embodied, and falsifiable alternatives.
What I said in the IAPCT talk was that prediction models (FEP in particular) are “… ‘red herrings’ that deflect attention from the search for the perceptual variables around which purposeful (control) behavior is organized.” So Madhur and I apparently agree that FEP is not only wrong, it gets in the way of doing what needs to be done to advance our theoretical understanding of how the brain works.
Best
Rick
Warren Mansell warren.mansell@curtin.edu.au
Sun, Oct 19, 2025, 11:29 PM
Thanks Rick, well said,
Warren
Madhur Mangalam mmangalam@unomaha.edu
Sun, Oct 19, 2025, 11:30 PM
Some more fun please!
Also,
Brian Fleming brianthebarkeologist@gmail.com
Mon, Oct 20, 2025, 9:23 AM
These articles are cathartic given the fatigue I feel from FEP hand-waving which seems to better serve my colleagues in feeling cerebral than it does in developing any testable predictions, parsimonious mechanisms, or god forbid, working technologies.
Most of my field (applied animal behavior) is still stuck in Behaviorism, yet the jump from Skinner to Friston seems natural for people who want to never feel theoretical risk & enjoy flexible retrofitting (as to Never Be Wrong & Never Contribute Anything).
Thank you for these hilarious & spicy papers!
-B. Taylor Fleming
(401) 218-8364
@BrianTheBarkeologist
Bruce Nevin bnhpct@gmail.com
Oct 20, 2025, 10:04 AM
Oh, this is so lovely! Fristothermopsycophantogenesis — You have me laughing out loud, Nadhur Mangalam, wow indeed!
As I said at our recent IAPCT conference, Friston’s nonchalance about verification & disproof satisfies the technical definition of bullshit: “Speech intended to persuade without regard for truth.” (Frankfurt 2005). Looking forward to reading these things and more.
On the other side, Henry Yin continues being brilliant, as usual. (Henry, I hope you don’t mind my adding you to this estimable group.)
https://www.nature.com/articles/s41467-025-64132-4
/Bruce
Is it true? Is it kind? Is it necessary?
Bruce Nevin bnhpct@gmail.com
Mon, Oct 20, 2025, 10:29 AM
In this 2018 interview, Friston recounts how at age 8 he watched small living control systems (woodlice) controlling to be in contained, dark spaces and to avoid some perceived attributes of brightly lit, open spaces. Looking right at control, he could not see it, and concluded that randomness is the secret of life, the universe, and everything. He says,
“Their scurrying had no purpose or intent: they were simply moving faster in the sun and slower in the shade. The beautiful simplicity (or nihilistic tautology) of this sort of explanation for life—and creatures like us—crystallized in my teens. It is exactly the same principle that underwrites the ensemble density dynamics of the free energy principle—and all its corollaries.”
A theory of behavior must be coupled with a theory of learning, and a theory of learning must include some kind of access to randomness as an ultimate resource — for PCT this is modeled as a reorganization system and, across generations, evolution. Friston, like other cognitive psychologists and like the behaviorists who they imagine they revolutionized into historical oblivion, promotes his theory of learning as a theory of behavior.
Theories that maintain Watson’s old promise, “prediction and control of behavior”, win more funding and political support than does a theory (PCT) which recognizes and explains autonomy and agency. Managers, commanding officers, and rulers aren’t interested in that. Collective control among autonomous control systems is hard to understand and hard to predict. They want the shortcut to making others predictable so they themselves can control better. How predictable!
/Bruce
Is it true? Is it kind? Is it necessary?
Madhur Mangalam mmangalam@unomaha.edu
Mon, Oct 20, 2025, 10:31 AM
Friston is a con man.
Madhur Mangalam mmangalam@unomaha.edu
Tue, Oct 21, 2025, 12:54 PM
Please find attached some published bibliographies on nonergodicity and Simpson’s paradox that did not struck any ears until now.
Btw,
Panphlogistodynamic Transmeta-Homeo-Self-Predictivity (PTHSP)
(noun; Fristo-psychpansy, ca. 2025)
A theoretical construct claiming to describe how every system, from neurons to nations, maintains itself by perpetually predicting its own self-prediction of the act of maintaining the prediction of itself. PTHSP is said to unify consciousness, metabolism, epistemology, and Wi-Fi connectivity under a single “variational umbrella,” whose spokes are allegedly “priors of priors of priors.”
Richard Marken rsmarken@gmail.com
Wed, Oct 22, 2025, 5:33 PM
Hi all
I think it was Madhur who said that Friston is a con man. But I don’t think he is. A con man knows what’s true and intentionally deceives his mark into believing what is not true. A magician cons (but in a fun way, unless he’s Uri Geller) because he intentionally deceives you into seeing something (eg., the Statue of Liberty disappearing) that he (and you) knows is not actually happening. I don’t think Friston is intentionally trying to deceive us into seeing FEP as what is actually happening when we see people behaving. I think he really believes FEP is the best theory of behavior there is.
I think Friston is more like a demagogue than a con man. He is convincing people of something he thinks is true, FEP, by taking advantage of the prejudices of his audience, in particular, the causal model of perception and behavior. Like other effective demagogues, such as Trump and Hitler, Friston is convinced that his belief is true because he shares his audience’s prejudice: Trump effectively convinced his audience that he won the 2020 election because he, like them, believed he was very popular; Hitler effectively convinced his audience that Jews should be eliminated because he, like them, believe they were the source of all the woes of society; and Friston effectively convinced his audience that FEP was true because he, like them, believed that behavior had to be based on inferences about the actual events that cause behavior. Oh, and like other effective demagogues, he puts on a good show.
But I don’t think Friston’s demagoguery is a problem. The problem is thinking that FEP is in some way “like” PCT. Or that it has something to contribute to PCT. It actually has nothing to contribute except obfuscation and distraction. Indeed, I would say this is true of any theory that claims to be similar to or that it could contribute to PCT. That’s because no other theory was developed to explain the controlling done by living systems. So no other theory makes predictions about the variables organisms control, how they control them and why.
Best, Rick
Madhur Mangalam mmangalam@unomaha.edu
Wed, Oct 22, 2025, 5:35 PM
I disagree. He is currently trying to get me fired for writing these articles, which only a con man would do. Please look into verses.ai, which is where the con is making money.
Madhur Mangalam
Oct 22, 2025, 5:52 PM
And of course, “there is no better theory than my theory.”
Warren Mansell warren.mansell@curtin.edu.au
Wed, Oct 22, 2025, 7:27 PM
I can think of multiple motives for wanting to get a person fired. Boris Hartman wanted me fired because he was convinced I plagiarised an idea he had about PCT which as far as I can tell involved simply reminding people that intrinsic systems have a closed loop internal to the body in addition to their effect on reorganisation. He believed both that he was right and that my inclusion of this idea amounted to plagiarism and that this was an offence that would substantiate being fired. I’m not saying your situation is the same as mine Madhur, but rather I think pluralism extends to reasoning about many situations!
Richard Marken rsmarken@gmail.com
Wed, Oct 22, 2025, 7:28 PM
Hi Madhur et al
On Wed, Oct 22, 2025 at 2:35 PM Madhur Mangalam mmangalam@unomaha.edu wrote:
I disagree. He is currently trying to get me fired for writing these articles, which onlya con man would do. Please look into verses.ai, which is where the con is making money.
Wow, I’m really sorry – and appalled – to hear that Friston is trying to get you fired. But we’ve got a demagogic President who likes to fire people as much or more than any con man. Nevertheless, you may be right about Friston being con man, though I don’t see what verses.ai has to do with it. Did Friston have something to do with it? Is there something fraudulent about it? Looks to me like just another AI system that happens to incorporate some Bayesian calculations.
Best, Rick
Tauseef Gulrez gtauseef@ieee.org
Wed, Oct 22, 2025, 11:50 PM
Verses.ai uses active inference and looks like Karl Friston has stakes in this project. It’s written in their research link.
Madhur Mangalam mmangalam@unomaha.edu
Wed, Oct 22, 2025, 11:59 PM
And here is the scam, for the doubting Thomasses:
https://royalsocietypublishing.org/doi/10.1098/rsif.2016.0616
Please compare the paper with the code implementation.
There is nothing called active inference, it is mumbo jumbo BS and Friston does not understand any math or programs. The guy knows only con. Plainclothesman know it all about Friston.
Warren Mansell warren.mansell@curtin.edu.au
Oct 23, 2025, 12:32 AM
Wow.
As far as I can tell, there’s no mention of the formulae regarding Bayesian statistics in the code? The code looks very brief given the extent of the maths in the article? Tauseef, what do you think of this, as you’ve inst[antia]ated a comparable model with PCT?
Madhur Mangalam mmangalam@unomaha.edu
Thu, Oct 23, 2025, 12:35 AM
FEP is the grift of the century and Friston the Charles Ponzi of neuroscience. Verses.ai is how they’re raking in millions from Saudi investors using a propoganda channel calles Machine Learning Street Talk. It is a pre-determined trajectory, that’s all. No prior, no nothing, pure Friston scam and has been all his careers even before FEP. Dynamic Causal Modeling was also a scam that did not go far.
Tauseef Gulrez gtauseef@ieee.org
Thu, Oct 23, 2025, 3:24 AM
Hi Warren,
The bayes is p = liklihood x prior / normalisaing factor, where as his minimum free energy is equivalent to max probability.
Min F = Max P
Also he performs Gradient descent on Free energy = prediction error x distance (mismatch)
Bayes is implemented through the gradient updates inside the DEM solver (as partial differentials of diif(ln P(x|y))).
I think where he does: DEM = ROS_spm_ADEM(DEM, postarget);
Best regards,
Tauseef Gulrez, Ph.D
Richard Kennaway richard@kennaway.org.uk
Thu, Oct 23, 2025, 5:36 AM
I greatly appreciate Madhur’s argued and forthright criticisms of the FEP school, and for drawing our attention to Verses.ai. I also found his “The myth of optimality in human movement science” (https://www.researchgate.net/publication/394497089_The_myth_of_optimality_in_human_movement_science) interesting, as I think it will be for all of us.
However, I do not condone the more extreme ranting about Friston being a con man, about scams and Ponzi schemes, and derogating his entire career, and even less Richard Marken’s rejoinder that he’s not a con man, but more like Trump or Hitler. (Have we all heard of Godwin’s Law?) The_Friston_Deprogramming_LOL.pdf is verging on green ink ravings. If I knew nothing about Friston and the FEP, my first impression of a document such as that would be unfavourable to its author.
This is not the way I would talk about anyone, even in the privacy of a closed discussion such as this one. And I am old enough that I do not need to think about defending my career. It is triumphalism in advance of a daydreamed triumph.
I mean, carry on as you please (you will anyway!), and I’m happy to be included in this address list. I’m just putting down a marker here, to distance myself from the more extreme words flying around.
(Per Vyv Huddy’s request, I have left his name off the recipients.)
– Richard
Rupert Young rupert@perceptualrobots.com
Oct 23, 2025, 5:57 AM
I concur. Very well said Richard!
Warren Mansell warren.mansell@curtin.edu.au
Oct 23, 2025, 6:54 AM
Hi Madhur,
I think you may have discovered the line to draw here, and it may help if you want to help people disengage from FEP without inflaming or personalising the issue.
We can still continue to benefit from your critiques of FEP, your clear arguments about the claims it erdocity, and your pluralist stance.
There is definitely a third way, not con man, not demagogue, but a persuasive, intelligent person who nonetheless takes science in a direction that the rest of us find distracting, counterproductive, overly reductive, and impractical.
I’m sure you don’t want advice on this, but one way to manage this would be to keep your academic arguments separate from the personalisation & satire, maybe with a pseudonym. But it’s a bit late now!
Talk to you soon,
Warren
Madhur Mangalam
Oct 23, 2025, 7:16 AM
Dear All ,
I won’t be continuing this conversation further at this time. I find it surprising that people who feel free to call a president of a country as great and important as the United States a con find it problematic to use the same term for a scientist—this is, frankly, hypocrisy. I believe this very reluctance to speak plainly is what allowed the FEP to reach its current status: too few scientists stood up, fearing that others might “cancel” them for doing so. It is deeply concerning that every discussion about the FEP seems to end in moral policing rather than scientific reflection.
I would also prefer no future communications on this matter. Thank you for your time. Please note that I take full responsibility for what I say, and I stand by it.
Best,
Madhur
Warren Mansell warren.mansell@curtin.edu.au
Thu, Oct 23, 2025, 7:27 AM
Hi Madhur, I think this conversation has got very confused and I’m sorry if those of you who’ve been circulated have found it frustrating.
I feel that I’ve found a fantastically helpful and insightful resource in your work.
Kind regards,
Warren
Richard Marken rsmarken@gmail.com
Fri, Oct 24, 2025, 12:54 AM
Hi Richard et al
On Thu, Oct 23, 2025 at 2:36 AM Richard Kennaway richard@kennaway.org.uk wrote:
However, I do not condone the more extreme ranting about Friston being a con man, about scams and Ponzi schemes, and derogating his entire career, and even less Richard Marken’s rejoinder that he’s not a con man, but more like Trump or Hitler.
I just got a chance to read your post and I apologize profusely for comparing Friston to Trump and Hitler. I certainly don’t think Friston has done anything nearly as bad as those two. My point was merely to say that I believe FEP is not a con (an intentional deception) but, rather, an honestly and firmly held (but unsubstantiated and probably false) belief. Perhaps T & H came immediately to mind as examples of people who held such beliefs because I was in Washington, DC at the time;-) But I can see that the use of T & H as examples was a terrible mistake and, again, I profoundly apologize. I only meant that Friston was “more like” those people in the sense of being ideological (believing without testing). But to prevent confusion I see that I should have picked more benign examples. Or, better yet, I just should have said nothing about it.
(Have we all heard of Godwin’s Law?)
I have now;-)
The_Friston_Deprogramming_LOL.pdf is verging on green ink ravings. If I knew nothing about Friston and the FEP, my first impression of a document such as that would be unfavourable to its author.
I agree. And I have nothing against Friston personally. It’s FEP to which I object, for the reasons that I gave in my IAPCT talk: It’s a red herring that deflects attention from the search for the perceptual variables around which purposeful (control) behavior is organized. The problem, of course, is that Friston is to FEP as Powers is to PCT or Einstein is to Relativity so it’s easy to conflate the man (or woman) with the theory.
Best regards,
Rick
Warren Mansell
Jan 5, 2026, 10:58 PM (4 days ago)
Hi everyone, Madhur has now blocked me from Linked IN, because I shared my perspectives based on PCT regarding his article. I was very polite and open for debate. This is what we are up against…
Warren Mansell
Richard Marken
Jan 6, 2026, 4:42 PM
Happy New Year Everyone.
On Mon, Jan 5, 2026 at 7:58 PM Warren Mansell warren.mansell@curtin.edu.au wrote:
Hi everyone, Madhur has now blocked me from Linked IN, because I shared my perspectives based on PCT regarding his article. I was very polite and open for debate. This is what we are up against…
I didn’t see the debate [on LinkedIn]. Can I find it somewhere? Maybe you could post an AI summary of it?
And whatever happened to IAPCT DIscourse being the place for discussion of all things PCT? Seems like all kinds of interesting discussions are going on all over the place except there.
Best, Rick