Really enjoyed chatting about a Drive to Survive with Carrie Figdor for the @newbooksnetwork.bsky.social and @mitpress.bsky.social podcast!

newbooksnetwork.com/kathryn-nave...

The aim of autopoiesis theory, as Humberto Maturana and Francisco Varela declared it in 1980, was to understand living systems via a ‘mechanistic’ approach that involved only what can be found ‘anywhere else in the physical world… blind material interactions governed by aimless physical laws’ (p. 74). In service of such a mission, in Maturana's words, ‘any attempt to characterise living systems with notions of purpose or function was doomed to fail’ (p. xiii). A lot can change in a few decades. By 2002 – to the presumed horror of those who had once rejoiced at this naturalistic ‘destruction of teleology’ (Beer, 1980) – we find Varela declaring that there is in fact ‘a real teleology implied in the notion of autopoiesis’, that it is a source of ‘subjectivity, intentionality and meaning’, and thus that, ‘organisms are subjects having purposes according to values encountered in the making of their living’ (Weber & Varela, 2002).

This search for an alternative ‘science of meaning’ is, alongside the rejection of a ‘representation-first’ view of cognition, one of the central pillars that defines Varela et al.'s (1991) presentation of the enactive approach. Yet the term ‘enactivism’ is also used more loosely to refer to the endorsement of the second, anti-representationalist, pillar alone. Disagreement thus persists about whether autopoiesis supplies an adequate basis for an alternative ‘science of meaning’, and indeed whether any such basis is even needed (Barandiaran, 2017; Ward et al., 2017). This broad tent of enactive cognitive science has been pitched across a fault line. On one side: cyberneticists, who hew to Maturana’s machinistic view within which mind may be continuous with life, but only to the extent that both are continuous with non-life and all can be subsumed within the mathematics of dynamical systems theory. On the other: ‘organicists’, who take living systems to constitute a genuinely new sort of organisation – one that is necessary for a system to be cognitive and which cannot be straightforwardly approached via the same modelling strategies used in ordinary physics.

New paper out in Adaptive Behavior!

On why cognitive scientists need a better account of purposive of behaviour, why cybernetics and dynamical systems won't do the trick, and how biological autonomy might.

journals.sagepub.com/doi/10.1177/...

BI 235 Romain Brette: The Brain, in Theory YouTube video by Brain Inspired

Interview with @braininspired.bsky.social for my book "The Brain, In Theory":

www.youtube.com/watch?v=T3zE...

‘The Brain, In Theory,’ an excerpt In his new book, Brette pushes back against theories that describe the brain as a “biological computer.” In this excerpt from Chapter 4, he challenges equating brain evolution with programming…

In his new book, @romainbrette.bsky.social pushes back against theories that describe the brain as a “biological computer.” In this excerpt, he challenges equating brain evolution with programming, and the universality of neural network models. #neuroskyence

www.thetransmitter.org/theoretical-...

Thanks!

‘The Brain, In Theory,’ an excerpt In his new book, Brette pushes back against theories that describe the brain as a “biological computer.” In this excerpt from Chapter 4, he challenges equating brain evolution with programming…

"The Brain, In Theory" is out today!

A short excerpt in The Transmitter @thetransmitter.bsky.social

www.thetransmitter.org/theoretical-...

Wanna do neuroscience in Paris but can't find interesting lab?

Want to come do a sabbatical but don't know who to collaborate?

Check this webpage aggregating ~all the neuroscience labs (+200) in Paris.

⚠️only the information of 'verified' profiles is reliable⚠️

Please retweet 🙏

parisneuro.fr

I wrote a somewhat critical piece on predictive coding theory and active inference.

"Predictive coding is not a theory of anticipation"

hal.science/hal-05565242

Reconsider the quote above: “an experience of pure darkness is what it is by differing […] from […] other possible experiences”. The claim is superficially appealing: we appreciate the particular quality of darkness in comparison with different experiences we have had before, such as lightness. But this works if we already had previous experiences. In other words, it presupposes that we are able to experience and to remember. One cannot define consciousness by presupposing both consciousness and memory. Thus, the fact that Integrated Information Theory is a panpsychist theory of consciousness is not an insight of the theory, but an assumption: the theory presupposes that there is a cognitive agent behind every state configuration, without explaining either cognition or agency. Both information by reference and information as difference define information as a potentiality: what an observer can infer from a signal. These are homuncular views, which take as prerequisite what they are supposed to explain, since inference requires prior knowledge (of external things, or of alternative possibilities). One cannot explain knowledge by assuming knowledge.

Just like cognitivism, Integrated Information Theory has a homoncular notion of information.

press.princeton.edu/books/paperb...

Excited to announce that Prof. Alicia Juarrero just joined the fantastic line-up of Keynote speakers for the "Beyond Neuro-Computationalism" workshop.
CFA here: www.uantwerpen.be/en/research-...
#philmind #philsci #philsky #philpsy #neurosky #neuroskyence

These phenomena have been termed “representational drift” or “dynamical representations”, but with the same logic, it could be claimed that the decimals of π represent the decimals of e, only with a representation that changes from digit to digit (Figure 5.7). It becomes rather unclear what is “representational” at all about such time-varying correspondences. Thus, the systematic analysis of the activity of neural populations has confirmed that “neural representations” are associations between experimental dimensions and neural activity that are stimulus-dependent, context-dependent, task-dependent, relevance-dependent and finally, time-dependent across various timescales, from milliseconds to seconds to days. In fact, the “represented” dimensions also depend on the dimensionality reduction method (Langdon et al., 2023). At this stage, the representational and coding terminology starts sounding like epicycles.

What is representational about drifting representations?

press.princeton.edu/books/paperb...

More good news: the special issue on “The Brain Abstracted” by the great Mazviita Chirimuuta just dropped @phimisci.bsky.social.

Check it out here:

philosophymindscience.org/index.php/ph...

(1/2)

Rethinking Life: Beyond the Machine Metaphor YouTube video by Marginalia Review of Books

I had a great talk with philosopher of biology Dan Nicholson, for the New Biology project with Marginalia Review of Books, about going beyond the machine metaphor in biology.
www.youtube.com/watch?v=0Ds0...

I wonder how Nature selects their comments. There must be dozens of people willing to publish a critical piece about AI in Nature, but somehow only the most uncritical, unrigorous praise gets through. Is that an editorial line? AI adoration only?

In Nature, you can write pretty much any kind of weakly argued tech propaganda you want, if it's a "comment" (which has become the majority of what was once a scientific journal). It's turning into a science tabloid.

Reality is not a controlled hallucination <p><em>The idea that 'reality is a controlled hallucination' has been recently popularised by figures such as philosopher Anil Seth. But this claim, which purports to be hard, down-to-earth science, i...

"Reality Is Not a Controlled Hallucination" by yours truly iai.tv/articles/rea...

Seriously I am stunned by how so many people seem to have abandoned the very idea of democracy. For example, AI-automatized peer review might be bad for science, but it's coming anyway so we have to adapt. What happened to the idea that people, not companies or robots, should be in control?

On the existential risks of artificial intelligence The impressive progresses in machine learning have revived the fear that humans might eventually be wiped out or enslaved by artificial superintelligences. This is hardly a new fear. For example, t…

The alignment problem is not between anthropomorphized AI and humans. It is between powerful profit-seeking organizations and the common good. Let's start worrying about real threats before we care about hypothetical threats.

(some thoughts here:
romainbrette.fr/on-the-exist...)

142. Dan Nicholson | What is Life? YouTube video by Friction

Here's another deep dive on 'What Is Life? Revisited' for Friction Philosophy.

Surely there's nothing more to be said about Schrödinger's famous book?

Think again!

Join us as we travel from quantum mechanics to genetics & molecular biology via statistical mechanics, thermodynamics, & cybernetics

Excellent summary.

In Greek mythology, Theseus managed to find his way out of the Minotaur’s labyrinth thanks to Ariadne, who gave him a thread he attached to the door. To escape the labyrinth, he just had to follow the thread back to the entrance. Out of the labyrinth, one could solve the path integration problem by tightening the thread, so that the thread takes the shape of the shortest path to the starting point, a straight line. Solving this problem involves no computation. Whatever Theseus’ position, the action that solves the problem is always the same: pulling on the thread. The procedure is neither computation nor information processing. Instead, it relies on interacting with the environment. In neuroscience and cognitive science, this is related to the concept of situated or embedded cognition. The interesting point about this example is that everything we have said earlier about the computational structure of behavior remains valid. It is still the case that the homing vector is obtained by iterating the operation x(t+dt)=x(t)+v(t)dt, and yet neither the brain nor the organism computes the homing vector. This is not just a question of boundary, i.e., of whether cognition occurs within the boundaries of the skull or is extended (Clark and Chalmers, 1998), because the thread does not implement the vector update operation either.

Finding one's way home can be solved by dead reckoning or by tightening a thread.
It follows that computation is not the only kind of problem-solving activity, even when the problem looks computational.

press.princeton.edu/books/paperb...

To build an electronic computer, stable states are constructed out of dynamical systems (flip-flop circuits), such that the dynamical nature of the underlying hardware can be entirely ignored: computation occurs as transitions between computational states, entirely shielded from the dynamics of electrons. Such shielding does not exist in brains. Thus, biological cognition cannot be reduced to elementary computations, supposedly implemented by neurons. Rather, computation is an elaborate form of cognition.

Computation is a particular kind of cognitive activity. It does not follow that cognition is entirely made of tiny computations (as cognitivism would make us believe).

press.princeton.edu/books/paperb...

Excited to share our new preprint exploring how Paramecium achieves diverse flow functions, i.e. feeding and swimming, simultaneously. This work was spearheaded by our ExM expert, PhD student Daphne Laan @daphnelaan.bsky.social :
www.biorxiv.org/content/10.6...

Great to see our paper on light-intensity dependent swimming patterns in #Chlamydomonas out now in Phys Rev Lett. as an Editors' suggestion! With a nice commentary by @philipcball.bsky.social.

Chlamy actively modulate the beat planes of their #cilia!
journals.aps.org/prl/abstract... #protistsonsky

nice!

Efficient coding theory postulates that neurons encode stimuli such as images in such a way as to transmit the maximum amount of information (Barlow, 1961; Olshausen and Field, 2004). This works by reducing redundancy, as in data compression. For example, in the Lempel-Ziv-Welch (LZW) algorithm, repeated patterns are assigned to shortcut symbols. In the same way, spikes are postulated to encode common patterns. This might sound like a reasonable principle, but the LZW algorithm is useful only because the table of correspondence between shortcut symbols and data patterns is also communicated, in addition to the compressed data. Without the table, one cannot decode the compressed data. Unfortunately, neurons do not communicate tables together with the encoded messages.

On the limitations of "efficient coding".

press.princeton.edu/books/paperb...

Yes, it's a circular explanation, exactly.

Well but the cell of that organism is supposed to be the product of the code, so it cannot be taken as a premise. The genomic code supposedly explains species differences, but you need the different species to already exist to express those differences.

More in this paper:
hal.science/hal-05491732
7/7

So what’s the genome then? The genome is a transmissible constraint on development, which channels the growth and division of a cell. The outcome always depends on what cell there was in the first place, as well as on other conditions of development. 6/7