Cyber Brain!

A reader recommended Andrew Pickering’s The Cybernetic Brain. This book is a survey of the British tradition in cybernetics, which it argues is a third, distinct strand of thinking that differs from the characteristic MIT and Soviet schools of thought. Where the US cyberneticians were inspired by engineering problems and saw it as a science of artificial intelligence, and the Soviets as a generalised theory of organisations, the Brits saw it as a means of inquiry into the sciences of the brain, and were especially inspired by biology and ecology, psychiatry, and fieldwork in sociology or in the practice of management consulting.

It’s also an impressive assortment of cool stuff, from robots that initiate a mating dance to Stafford Beer’s putatively intelligent pond just outside Sheffield. (Don’t you want a science-fiction alternate history where that thing worked?) Apparently, Beer and David Suzuki put together a high-concept 80s coffee table photo survey of the history of computing, including Japanese fishermen’s rope tallies and SGI workstations. Don’t you want one of those? No? What are you doing on this blog?

More seriously, the lesson that stuck out for me was that a lot of the work in neural networks, expert systems, and embodied cognition was rediscovered since the 80s in robotics and in theoretical computer science, and since then a hell of a lot of it has found applications. I suspect that the challenge now may be to get the emerging disciplines of user-experience design and service design to rediscover it in their turn.

To some extent, the cybernetic tradition of hardware-hacking and rapid prototyping as a method of inquiry has stuck around and prospered with the UX people, but I’m not sure if the critical element has got across. Beer didn’t just ask if this worked, he asked if it should work. The huge recommendations systems of the e-commerce majors are a fine example of the kind of artefacts cybernetic ideas were intended to help us explore and learn to live with.

Things I didn’t like about the book: the weird framing of the Iraq War as a big modernist planning scheme.

The individuals who drove the project on were very much informed by critiques of planning, by complex-systems studies, and by cybernetic and systems dynamics ideas. They were libertarians and people who spent their careers building the network-centric warfare systems of the 1980s! That’s what “regime change” means – a lot of people assumed it simply meant “replace the government of Iraq with a different one” or even “change some features of the government of Iraq”, but in fact it means “cause the whole Middle Eastern political system to flip to a different state”.

Rumsfeld had his “regime change”, and Petraeus’ “complex-adaptive systems approach” was called for to clean up the consequences. On a less intellectual level, all sorts of Iraq boosters constantly attacked experts, Arabists, and the very idea that a plan might have been useful.

Generally, I wasn’t totally convinced by Pickering’s philosophical distinction between “performative” and “representational” knowledge either. I understand the point, but it is perfectly possible to have a working, emotional/kinaesthetic understanding of something that is also horribly, disastrously wrong, and inhumane to boot. The history of medicine is a great source here.

Speaking of which, boy, there’s some deeply creepy mid-century psychiatry in here. Ross Ashby, who coined the phrase “requisite variety”, also sought a “Blitzkrieg method” of treating his patients that might include “hypnosis, then LSD, and then ECT”. William Burroughs, who repeatedly pops up in the book as a voracious reader on the subject, would surely agree that “gentle reader, this prospect buggers the imagination.”

6 Comments on "Cyber Brain!"

  1. I was more interested in the historical aspects of the Pickering book than the more philosophical bit.
    Pickering also wrote a good and quite long paper on Ashby, which is probably as close to a biography as he is going to get. I can’t remember if the chapter on Ashby is essentially the same material or not – will check at some point. The LSD + hypnosis + ECT is truly mind boggling, but so is the idea of a private mental hospital in Gloucestershire having a research director – it would be interesting to have more context for how out of the ordinary this was.


  2. Unfortunately I am one of those readers of this blog who knows nothing about robotics etc, so could you make it clearer what is Pickering’s framing of the Iraq War and what is your critique of it?


    1. As a high modernist planning project in James C. Scott’s sense. My critique is that the leaders of the Iraq war were influenced even more by anti-planning and post-modernist ideas, drawn from economic libertarianism, mission command, MIT-style systems dynamics, John Boyd’s OODA loop, and precisely cybernetics. “Regime change” refers to an event where a system transitions from one stable operating regime into another.

      Rumsfeld, Shooter Dick, Wolfie et al specifically didn’t want a plan and didn’t want experts, because they wanted to inject energy into the system and let the pieces fall where they may, imagining that as the initiators, they would also be the first to react to the consequences. Remember the quote about “we’re an empire now and when we act we create our own reality”? Better known for the mockery of the “reality-based community”, but it exposed a whole philosophy of creating alternative truths by action.


  3. If the neo-cons really thought that they could change a large system by cybernetics, then they could at least have read chapter 1 of Ashby’s ‘An Introduction to Cybernetics’, where he says in s.1/5, “In this discussion, questions of energy play no part – the energy is simply taken for granted”.

    If you get into such activities you need to have infinitely deep pockets, otherwise you can end up playing the Martingale like Dan Bernoulli’s girlfriend. The finite size of your defence budget acts as the casino limit.

    Ashby’s Cybernetics allow you to set up a lot of neat state-transitions. When you deal with biology, and understand the what replicators are out there, you can see how the state space gets explored. What is missing is purpose and optimality. Biological systems do not have purpose in the way systems engineers talk about. Evolved systems are not optimal.

    If you have a purpose, a budget, and are concerned with optimality then exploring complex systems outside a simulator with (other bad people’s) real money is a mugs game. This is what they teach you in system engineering courses. Of course when you go into business the budget bit gets hoisted up a level to ‘adult supervision’. Possibly this is why few large engineering projects run to schedule, budget and quality in Blighty.

    Oh yes, Operation Cast Lead.


  4. That’s what “regime change” means – a lot of people assumed it simply meant “replace the government of Iraq with a different one” or even “change some features of the government of Iraq”, but in fact it means “cause the whole Middle Eastern political system to flip to a different state”.

    Huh. That makes an awful lot of sense, suddenly.


    1. Yes, this. Also Alex’s comment at January 3, 2014 – 12:22 pm really puts it in sharp perspective for me.


Leave a Reply

Your email address will not be published.

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