Markov Blankets vs Lipid Membranes

Watching this wonderful “Krebs Cycle” RI lecture by Nick Lane – last mentioned him here – and just wanted to note two things for now.

One, there is an obvious topological parallel (*) between (System) Markov Blankets and (Mitochondria) Lipid Membranes at two quite different levels of abstraction – information processes and biochemical processes.

And two, early on he makes quite a few remarks against the informational-computational view and yet highlights the information element when we get to the nucleotide processes.

Fascinating for the metaphysical aspects of which came first – which is of course what his talk is about, the primacy and universality of Krebs Cycles in anything we’d recognise as biological life – from the simplest physio-chemical precursors to the most complex multi-celled creatures.

AND the inevitability aspect that once the simplest “chemistry” exists evolution of the complex follows.

(Lots to unpick on further detailed review – but fascinating to note the above on first pass. Also lots of good acknowledgments of the women involved in the research processes.)

(*) He even mentions the topological parallel at the whole earth level!!!

Q&A here too:

Oooh! and a Jeremy England question mid-way through the Q&A.

LUCA – Last Universal Common Ancestor of both the archaea / bacteria / mitochondria and the eukaryote cells – not unlikely coincidence – everything to so with structure not info – yes, yes, yes system architecture level info, not individual bits.

Wow! It’s all there.


[Post Notes:

Watching this Luis Razo & Karl Friston talk, extending the topological parallels – beyond Razo’s balloon – even the event horizon of a black hole is a Markov Blanket. Every “thing” to which you associate individual states has a Markov Blanket in “phase space” – whether there is a physical membrane or not. All you need to know about that thing is at (or projected onto) that surface. As he says this is about the definition of using the word thing in any sentence. (Hat tip Active Inference folks.

This theory of “things” fits well to my “distinguishing between A & B” diagram in this “Identity” post.

And Razo’s bio-electricity … significant in the Nick Lane presentation subject of this post! It’s perfectly OK to capture semantics in relations involving the thing, but let’s not lose the thing itself. Curiously – I note Nick Lane is referenced by Mark Solms but Mike Levin is not. Levin is however referenced by Dan Dennett – Tufts connection. Lots in here. Thing-ness is the key – see next post re modelling tools.

“The Everything Crisis” – Globalisation – yes. The problem is the destruction of Markov Blankets, every time we lose a boundary we lose some “thing”. We are definitely on the same page.]

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.