Life 3.0

Max Tegmark’s full title is …

“Life 3.0 – Being Human in the Age of AI”

… and I remain pretty sceptical about the AI-hype. Last mentioned Max’s book back in 2017, but already added a 2019 note to that just last month.

The book, as the Life 3.0 title suggests, is primarily about evolution and, as director of the “Future Life Institute”, he takes his layered model into futurism – predictions about the future. As I say, I’m sceptical about AI ever getting beyond automated algorithms until such time as “real” A-Life has also evolved, and I’m sceptical about predictions generally. However, his layered 1.0 / 2.0 / 3.0 model is based entirely on information patterns AND he has an entire chapter “Goals” on the cosmic teleology implied in the inevitability of the 2nd Law statistically favouring increasingly intelligent life.

This cosmic inevitability of intelligent life, human and beyond often attracts “anthropic” criticisms, but such criticisms are political, the political correctness of orthodox science that insists humanity ain’t that special. This was the life’s work of Rick Ryals about whom I posted a page only last week. Rick was primarily a physicist with cosmological interests, as is Tegmark though for him the mathematics is the real focus. He uses maths where I say “information”.

Actually, despite my scepticism, the whole book is a good read, well written and witty in his everyday examples. A red-pill / blue-pill matrix allusion in the idea of the AI singularity overtaking humanity, plenty of good thought experiments. As ever, I’m just gutting the book to satisfy my focus.

Given my anthropic-information / layered-evolution focus, Tegmark’s story seems entirely consistent. Encouraging to see Jeremy England referred to as a colleague at the FLI, since it was England’s work that Dan Brown referenced explicitly in “Origin”. In fact Tegmark could be as much a candidate as Ray Kurzweil for the protagonist in that work.

Everything really is hanging together increasingly consistently, reinforcing the impression that I need to shift my focus from reading to writing yet again.

Life Versus Entropy

Prompted to post this morning thanks to the click-bait statement “Schrödinger spread the misconception that life works against entropy” in this tweet:

In fact it’s a link to an Arxiv paper by Kate Jeffery and Carlo Rovelli, one which I’ve had bookmarked for a couple of weeks. I think Carlo shared it when it was published.

On the statistical mechanics of life: Schrödinger revisited

Along with Carlo Rovelli’s

Where Was Past Low Entropy?

My take is that Schrödinger is not misleading provided you make the distinction between local and global effects. Schrödinger is definitely worth revisiting in  order to understand, but he’s not wrong. In fact life’s “efficient” drive to replace entropy with order and excess energy (under local exploitation) is part of the drive in the global entropy increase of the cosmos. Clumping of material structures in vortices and galaxies (ie at all scales) is part of the same drive. Bacteria and mould in your compost heap. Local order, globally accelerated chaos. Life is “simply” the next stage after bio-chemistry has evolved from physics.

That statement is a summary of the position I arrived at thanks to Rick Ryals (who sadly passed away Dec 1st, 2018).

The drive is a kind of teleology, one that makes increasingly complex and intelligent life “inevitable” in an evolving cosmos, one evolving to its maximum entropy end according to the 2nd law.

I last mentioned this stuff only a week or so ago, in the footnote here “Who’s In Charge?” . Also I need to posthumously archive Rick Ryals “Entropic Anthropic Principle” material and have had Rick’s original 2006 “post” bookmarked recently whilst I cogitated.

‘So the second law of thermodynamics is never violated when the entropy of the universe always increases via the described perpetually inherent thermodynamic structuring, which enables the universe to continuously “evolve”.’

Also, this quote from the Rovelli / Jeffery paper:

[L]ife is not an improbable “fight against entropy” —as Erwin Schrödinger famously put it in his adventure into biology (Schrödinger 1944)— but is rather a statistically favored process directly driven by entropy growth, in which movement of a system within a space of available states leads it to discover —and traverse — channels between metastable states.

Absolutely – “a statistically favored process directly driven by entropy growth”. Life builds order and excess energy locally and temporarily, but is “driven” by overall long-run entropy growth. It’s “statistically favoured” – so is an inevitable part of the story.

Also significant, the concept of a continuously evolving universe. “Evolve” in scare quotes. Another current scientist who seems to get the structuring at all scales is Erik Verlinde, who also looks at entropy as the inverse of information, fundamentally limited in its density at the event-horizon of black holes. A topic on which Sabine Hossenfelder posted just last week, and dissed Darwin in the process!

That information density limit is the true quantum of fundamental physics. (See Smolin’s view-based realism.)

It’s all happening.

Science and the Man

Been checking my previous references to “Epstein” but none seem to be the “Jeffrey Epstein”. I checked because it’s quite clear that the (convicted and now dead) billionaire who liked to party and throw his money around to win friends and influence people is clearly “connected” with many science types through Harvard / MIT / Santa-Fe and the Edge publisher John Brockman.

The creepy Larry Krauss is one I’ve mentioned a few times before. I long since decided he was a charlatan science-wise, since it was always clear he’d prefer to ignore difficult questions that might undermine the main thesis of his current book and lecture tour sales. Now, in philosophy or science, getting books published and marketed to best-selling proportions is a tough -messy, commercial – business that goes well beyond their content. I’m not one of those who gets sniffy at academics needing to get commercial reward from book sales. Sexy sells, figuratively speaking at least. Krauss and Dawkins and Hawking are simply examples of where the successful marketing messages get hopelessly confused with good science, to the detriment of real science (and philosophy and public knowledge) as a whole. Most of it concerns the politics of getting their next project funded.

My own main agenda, really, is to counter this with concerns for the supposed self-correcting aspects of the methods of science. (They’re way out of kilter with the speed of modern communication channels systematically non-robust, beyond-fragile, technically non-anti-fragile.)

Frankly I have no idea what Epstein’s real interests (or credentials) ever were in science, but it’s clear he’s funded a few scientists and their projects along the way. The billionaire gets the power buzz, the participants get the attention buzz. A proven recipe. Unless Epstein had some almighty “sting” project in mind I can’t imagine what other motives there could beyond the power-play dependency with the added buzz of illicit sexual danger.

I’m no conspiracy theorist, and have no idea how many of the academic types were where on the knowing / reluctant / willing scale, how many actually participated in the available options or how many had their scientific agendas distorted by their motivations. 90% of science is crap, because 90% of everything is crap. Cock-up rules. If you’re Brockman, money, connections and publicity can only be good for marketing and sales.

The whole thing is very sad, not least for the underage victims criminally exploited. But defending the alleged actions – a la Krauss – is far worse, far beyond regret at errors of judgement.

I still find Brockman’s Edge a useful resource. (Most recently from 2017 unedited) (Many more going back to 2004) Even if only 10% of the objective content is any good, that’s valuable if you have hundreds of individual resources and you can smell the quality. In selecting nuggets from answers to the annual question (4 of the 14 in 2017 were women), I’ve remarked before that part of it is recognising the career politics of the contributors – is this a career-defining idea they need to sell, or are they comfortably into the tenure phase of their first careers or are they onto their second and third careers and can simply afford to be speculatively mischievous? The “business models” used in academia are clearly a root problem here.

Science and the man are inseparable.
A lot of it stinks.


Hat tip to Sophie Scott for sharing this tweet from Rebecca Watson, and to the whole thread of linked resources:


Dan Dennett has long been a hero of mine, the science-friendly philosopher who has synthesised both Darwin and Turing into his theories on the evolution of conscious will. Great 30 minute interview here with Jim AlKhalili on his “Life Scientific” on BBC Radio 4.

A potential new hero here in John C Doyle of Caltech, someone I’d not been conscious of until I saw him referenced by Michael Gazzaniga (the last few posts). Not managed to find any on-line (eg social-media) contact other than his Caltech email (so far no response) nor any writings other than his (supposedly out-of-date?) teaching materials, but he claims here to have many on-line videos. In this particular YouTube example from a 2018 conference, one of his first slides features both Darwin and Turing. So, another man after mine own.

An excellent presentation on Universal Architectures enabling evolving optimality of system functions and performance – from information thru biology to mountain-biking(!) And more: Integration of direct (visual) and indirect (propriocentric) network pathways – crucial architectural role in evolution of key operating systems despite multiplicity and flexibility in the range  hardware and software options – layered in many domains (hence universal). Human complexity out of many layered networked algorithmic simplicity with many orders different comms response speed in the different layers / channels. His “robust-yet-fragile” tag (a) suggests Taleb and (b) suggests a genuine homeostatic balance – a dynamic sweet-spot.

Viral fragility of the OS in such architectures. Bad Memes – OS Infections – are humanity’s biggest problem. Solution is moderation, slowing down the response times of the network layers / channels susceptible to such viral memes. (I could have written this stuff myself, in fact I have many times!) Even back in 2007 Doyle was warning the internet needed controlling – moderating as I’ve been calling it.

Everything except any questions of causation (supervenience and/or independence) between those layers. Intriguingly, determinism is one topic mentioned on several of his slides, often qualified “worst-case”, but not highlighted in anything he says. (All his published stuff looks like pure control system theory text-books – robust yet fragile.) Searching on.

[Clearly he is “very famous” in control theory – eg for his 1978(!) paper demonstrating arbitrary fragility (lack-of-robustness) in control systems with uncertain time-delays in certain measurement and response loops > leading to Robust Control Theory.]

Who’s In Charge?

Rounding up my reading of Michael Gazzaniga, his 2015 “Tales From Both Sides of the Brain – A Life In Neuroscience.” in particular, though having completed and enjoyed it I went back to his “Who’s In Charge” for a second go. Previously, newest first:

Who’s In Charge (WIC) first: I went back to after initial doubts because Tales From Both Sides turned-out pretty good. I see now it is the transcripts of his 2009 Gifford Lectures, so tailored to a particular kind of message for a particular audience. His sub-title is “Free Will and the Science of the Brain” and he gives a good overview and summary of the known science and positions, but ultimately is non-committal on what he really believes. The chapter “Abandoning the Concept of Free Will” is disappointing because I’m pretty sure he hasn’t abandoned it, he’s just providing the orthodox science story on why scientific determinism says “it can’t be real” despite our (his and my) strong sense of responsibility for actions in ourselves and others.

It’s much clearer in TFBS, the later book which, despite being essentially an autobiography, is much more bullish on what he really believes about the science – putting his money where his mouth is.

[What follows is more notes than a review
– it sparked off so many linked avenues

Free will is real all right, we just need to upgrade our take on what causation means to the deterministic science of reality. As I’ve said before “Super-Determinism Sucks”. At the beginning and end of the offending chapter he makes a reference to John Doyle and his hardware<>software systems approach, but doesn’t elaborate nearly as much as he does in the later book.

Doyle is (I think) new to me, but he clearly holds pretty much the same many-layered complex adaptive system view as I do on why real conscious agency evolves in the higher layers. The evolutionary view is crucial to both the brain and mind stories, hence Dennett, hence EES.

TFBS emphasises that the core debate in brain-mind science is about supervenience vs supersession between layers. WIC doesn’t even mention them. Talking simply in terms of physical and mental as two layers as opposed to the real multiplicity I call onion-skins, Gazzaniga quotes Donald Davidson:

“Supervenience might be taken to mean that there cannot be two events alike in all physical respects but differing in some mental respects, or, that an object cannot alter in some mental respects without altering in some physical respects.”

Quoting Sperry, Gazzaniga’s one-time boss, he refers to lower layers being outclassed or superseded by higher layers:

“where level-n floats freer than level-n-1″

Although as he puts it:

“unrepentant reductionists see a sleight of hand here.”

Super-determinism sucks, as I say.

Although I’ve given up in recent years being overly precious about the definitions of supervenience and supercession, the important question is whether an evolved state in a lower layer uniquely determines the evolved state in a higher layer or whether state and causal relations between states in the higher layer can be in any way independent or free of a given state in a lower layer. If the latter, clearly interesting questions of identifying “cause” arise in the higher layers. As I say precise definitions here, still boil down to what we mean by cause, and this is very much Doyle’s point.

As he goes on to say, Sperry is talking about causation being something more than neuronal firing. As Gazzaniga has already elaborated at this point, there is something like downward causation, although he rejects calling it that and, wrongly in my view, brings in a sort of quantum complementarity and uncertainty for some kind of indirect causation (See footnote).

Long story short, there is some kind of causation in higher layers that is independent of lower layers. It’s about evolutionary and current time-scales – learning and action are circularly related. Also as Gazzaniga shows there are many cueing relations beyond communication by neuronal firing that contribute to cognitive and motor processes – somatic and propriocentric. Decision-making loops pass through many of these layers and many are gamed between anticipatory and subconscious guesses and both subconscious and conscious reactions to new information (hence why Libet is wrong, he agrees).

After all this, in WIC, Gazzaniga still makes the illusory free-will claim that it’s just what our “interpreter” left-right brain integration wants us to think. (For me this suffers the infinite regressive humunculus problem, that for the interpreter to “claim” it needs will to cause. Again as I’ve said before this regression is not infinite, it resolves into layers. I think I may have to read the original Doyle?)

Towards the end of the final “Layers and Dynamics, Seeking New Perspectives” chapter of TFBS he gets very close to saying that the evolutionary engineering take on layers of causation in a complex adaptive system that provides the agency in the mental layers. Layers note, not layer.

The left-right cortex, corpus-callosum and sub-cortical distinctions – which fill the majority of the content of TFBS – clearly support the permissive supervisory control aspect of a many-layered complex adaptive system. Even if the willful control is itself a complex problem of guessing and gaming, feed-forward and feed-back, there is no mystery – for me. No quantum weirdness necessary.

Very interestingly Gazzaniga, just before winding down from his story so far at this point makes reference to “very recent work” by Giulio Tononi. He also makes positive references about work at Santa Fe. I’m already there with Extended Evolutionary Synthesis (EES) and Integrated Information Theory (IIT). Doubly interesting, I picked-up Tononi from Chalmers who I think was largely responsible for so much of the hard-problem, subjective-agency and supervenience doubts. Wonder if he’s progressed too?)

So, do I read the even later Gazzaniga in his 2017 The Consciousness Instinct: Unraveling the Mystery of How the Brain Makes the Mind or do I find some original Doyle? Has 2017 Gazzaniga progressed since 2015?


[Footnote : Note also, Damasio is part of Gazzaniga’s story and is part of this Dutch session on “feelings”. He makes a big thing of life as homeostasis – working to achieve active energy surplus to re-invest in flourishing, as opposed to some simple equilibrium or static balance to “stay alive” – which is death of life of course. He also makes a big deal of the layering conscious / subconsciousness, cortical / sub-cortical in the mechanisms of feelings arising and being felt consciously as a state, a “quality of life”. The first question after Damasio’s talk is Q – “is that energy of homeostasis the surplus needed to counter entropy?” A- “Yes.” Asked by Eddo Rats, a friend who shares the multi-layered complex-adaptive-systems evolved-engineering view – yet still holds the need for Penrose-Hammeroff quantum coherence to explain the “downward causation” of mental-agency. For me it’s much simpler and the “entropy” is the clue. Information – an entropy-complementarity – is more fundamental than both physical and mental – hence IIT and relational-information – a pre-conceptual quality – metaphysics. Eddo’s interest is more psychological and psychiatric than metaphysical, but for psybertron there is no difference.]

[Post Note: Researching John Doyle I find this Caltech page, but no general readership book publication, and no obviously directly relevant papers to which Gazzaniga might be referring. And, there are no specific bibliographic references in Gazzaniga?]

Mariana Mazzucato

Bought her latest book “The Value of Everything” on the strength of an excellent short-section of her talk here:

As I said at the time, I love the style as well as the content, with discursive parentheticals and “air-quotes”. A great communicator.

Picking this up before a final review of my foray into Michael Gazzaniga couldn’t help notice the syllabic similarity in their names. But more significantly, early on, we discover that one of her mentors was / is Carlotta Perez of Freeman & Perez fame during my MBA Economics days. A lot of influential women in this story. Economics needs women.

As you’d expect lots, early on, on the price / value confusion and the objective / subjective confusion in distinguishing the one over the other. Real issue is about economic models allowed to exploit the confusion to counter-productive (low value) ends, rather than definitive arguments. Too soon to say, but on the right track. Reading on.

Baloney Generator #2 – The Most Stunning Result.

I first mentioned “the baloney generator” back in May 2003 – actually in this hand-crafted-html (!) review of Steven Pinker’s Blank Slate in December 2002.

“The conscious mind — the self or soul — is a spin doctor, not the commander in chief.”

The accusation being that the rational mind generates narratives – any old baloney – to explain incomplete or ambiguous states of knowledge, so that it can move on. It’s controlling the (explicit) news to get on with it’s own (implicit) agenda – like McGilchrist’s “Berlusconi” metaphor.

Of course the 2003 Chicago Uni link is long dead, and anyway from the references above it’s clear the concept was already well recognised by others even if it was new to me at that time. I also mentioned Pirsig’s Lila in the same Chicago context(!).

Now it’s not fashionable to refer to Pinker these days, he’s not stupid but he’s maybe not as smart as the scientific assuredness of his various claims might suggest. None of us is perfect, not even Taleb 😉

Anyway, not only had I forgotten Pinker as my source of baloney, I’d forgotten he had been a colleague of and had written the foreword to the Gazzaniga that I’m currently reading. I say that because I found myself scribbling “baloney generator” in the margin.

Describing one of many split-brain subject experimental procedures – behavioural responses to left-right, eye-brain-hand-arm stimuli and movements – he writes:

[Instead of asking] “WHAT did you see?” [A] “Nothing / no idea.”
[We asked the subject] “WHY did you do what you just did?”

In simply changing the question a virtual torrent of new information and insight flowed. Though the left hemisphere had no clue [it was disconnected from any signals] it would not be satisfied to state that it did not know [why]. It would guess, prevaricate, rationalize and look for cause and effect, but it would always come up with an answer that fit the circumstances.

In [Gazzaniga’s] opinion, it is the most stunning result from split-brain research.

When it comes to reasoning, thanks to the left-brain, the mind can be (is designed to be) a baloney generator.


[Post Note : and how could I forget, the reason I was prompted to post was because I was asked to name philosophical fiction other than Pirsig or Gaarder that I considered important. Anna Karenina was my surprise suggestion, but Crime & Punishment is more obvious. The relevance? Razumikin’s rant:

“Talking nonsense is the sole privilege mankind possesses over the other organisms. It’s by talking nonsense that one gets to the truth !

I talk nonsense, therefore I am human.”

It is talking baloney that makes us human, if there are any logical positivists listening?

“Talk nonsense to me by all means,
but do it with your own brain … “

It was even a Pirsig conversation that led me to that particular Dostoevsky!]