AI in the Academy: Unfolding dynamics, and a history lesson

There is a mystery as to why the most transdisciplinary science, cybernetics, never really got a hold in the university. Yes, there were weird outposts of cybernetic activity like Von Foerster's Biological Computer Lab at the University of Illinois, but it turned out to be not very sustainable. The most significant major UK centre was in Hull University, and that has pretty much been disbanded. I believe what we are seeing with the impact of AI in the university is telling us why this happened, and why a similar pathology is happening again. 

A university is a set of disciplinary fiefdoms - elegantly described years ago by Tony Bucher's "Academic Tribes and Territories". Academic tribes or fiefdom's tend to want to defend themselves from each other. When disciplinary boundaries are clear, this works pretty well - and has done since the trivium and quadrivium of the middle ages... 

When a truly transdisciplinary subject comes along - and cybernetics was just that - it puts disciplines in a bit of a panic. It's not that they want to defend themselves from the transdiscipline, but rather they each seek to own it, and therefore look to acquire and colonise bits of it. We are seeing exactly this process unfolding around AI at the moment: every discipline is staking its claim to AI. The consequence of this is that the transdiscipline becomes divided and absorbed into disciplines. Its intrinsic transdisciplinary nature is dissolved. This is truly crazy behaviour, but it is determined by the structure of institutions. 

The only hope I feel is for universities (or some other institution for scientific inquiry and intellectual growth) to construct themselves not around the codifications of curriculum and disciplines, but to construct themselves around tacit knowledge, shared experience and creative expression. AI could help here. It could be a huge amplification of the creative imagination. It could create shared experiences in ways we have not conceived of before. Unfortunately, despite their many virtues, universities are unlikely to be the home for these new kinds of innovations and experiences. It will happen somewhere else. 

The issue has to do with the roots of institutions of knowledge and science in monasticism. Whatever caused human beings to retreat from daily ordinary to live in ascetic small communities in the desert was driven by a deep physiological need. Many scholars feel this same need in the wake of the modern academy's transformation into a business. Physical retreat is unlikely. But a spiritual retreat to new kinds of shared experience and ways of communicating is likely to become more feasible. Students and industry may follow, and the universities may have to play catch-up. 

Territoires de l'oubli - Spectralism, Difference and Paraconsistency

I went to a concert in the Barbican a couple of weeks ago where the BBC SO played Tristan Murail's Gondwana. There were a number of other hybrid electronic/acoustic works. As a now-elder exponent of musical spectralism (where audio analysis of sound waves themselves informs dynamics, instrumentation and form), I found the Murail piece particularly fascinating. Perhaps more impressive than Gondwana, is his piano music - particularly this piece "Territoires de l'oubli":

Perhaps not the sort of thing one would want to listen to all the time, but it is fascinating because of the way the thumps and bangs resonate with the higher compass of the piano. I find it like listeneing to nature - rather in the same way as Messaien (who of course is the composer who springs to mind in listening to this). But this isn't about birdsong or God. It's fundamentally about physics. 

I also think it's about difference. The art of Murail is an art of identifying small differences in the detail of a waveform, and reproducing these differences in instructions to players such that the reproduction of small differences is almost as imperceptible as they are in our auditory experience.

What is a "small difference"? It is clearly related to what Ernst Weber called a "just noticeable difference" - an idea which became the foundation of psychophysics. It subsequently became embedded as a principle in an AI innovation I was responsible for, so this is a bit personal for me. There are big philosophical questions around psychophysical principles because they appear to run against the Kantian notion that there is an unknowable thing-in-itself about which we can have no knowledge. Weber, and later Fechner, suggested that there were indeed empirical things we can do - and Murail is clearly experimenting with these.

A small difference is liminal - existing in the space between presence and absence. It is the beginning of where what Graham Priest calls paraconsistency (where true and false co-exist simultaneously) arises (Graham Priest - 4. What is paraconsistent logic?)

Increasingly it seems that paraconsistency is the only useful logical position to take in our incredibly tortured world at the moment. Warren McCulloch, in his first paper on neural networks in 1942, articulated something like this and prefigured a paraconsistent logic in discussing the circularity of the nervous system (A heterarchy of values determined by the topology of nervous nets | Bulletin of Mathematical Biology). A "heterarchy of values" is a paraconsistent logic.  

This isn't a new idea. It goes back to the idea of "synchronic contingency" which was a key feature of the philosophy of John Duns Scotus. That it has acquired new resonance with quantum mechanics is perhaps an indication that the medieval insights were correct. I'm revisiting some of this stuff: Improvisation Blog: Information and Syncretism: from Floridi to Piaget

Our distinction-making is never objective, yet it result in selections from possibilities where we take what we select to constitute "reality". The logic of selection is not Aristotelian, as McCulloch argued. Yet we take it to be so - particularly from an organisational perspective. This is what is producing the aporia that see unfold in the Trump/Russia/Ukraine pathology. If we examined this from a paraconsistent perspective, how would it look different?

Rather like Murail's music seeks to unite the form of music with the physics of music, what if we united the form of decision with the biology of decision? Could AI, which is also a paraconsistent technology, help us?

Russia

There is no other productive way of thinking than to think of ourselves as a species, rather than as nations, corporations, individuals or political parties. We are a sick species and the only resource we have for healing ourselves is our mind. Unfortunately, our sickness is a sickness of the mind. Can a sick mind heal itself?

Beer on Transduction

 

Beer's Tribute to Laws of Form

 

The long time of emotional inertia

The feeling of a long time passing seems to be a feeling of a kind of emotional inertia: a time where there is a prolonged sense of unfulfillment, absence or expectancy. My friend Peter Rowlands says that the measurement of gravity is really the measurement of inertia. That's a profound statement: we perceive gravity by the force exerted on our head by a falling apple. In the same way we measure time by measuring space. A long time is a slow journey over a long distance. 

Emotional inertia is how we perceive "a long time". When we feel no emotional inertia, time passes very quickly. That's not to say that we can't be happy in a time of emotional inertia, but it is a different kind of feeling. 

After Trump's inauguration we are beginning "a long time". I'd rather not watch the news in that time. But I'll keep an eye on the wars...

Emotional inertia is a kind of unstable stasis - does that make sense? What is impacted is the true expression of self. The self can't express itself, but it still refers to itself. But it hasn't gone. It's just waiting. That's partly because the experience of inertia is our construction of our limited perception of the universe. The universe itself knows no inertia and no time. 

West Side Story

It's been a ridiculously busy week this week - no time to think. But I went to the Spielberg West Side Story in Wythenshawe this evening - part of an initiative to bring arts to Wythenshawe. I think Spielberg's remake is a masterpiece. I first saw it on a plane to China. Embarrassing because it made me cry - always difficult to say you want another coffee when that happens!

What's so extraordinary about this version is the sensitivity with which the love scenes are played. The timing is exquisite - Spielberg's musicality is not just in his use of the music, but in the choreography of the camera and pacing of gestures. But it just tells us how music is at the root of emotion. Everyone relates to this because everyone knows how this feels - even if not everyone actually experiences something so intense. But some people do. 

But the process of making something so sensitive is not something that is driven by feeling, but by intelligence and technique. One of the other things that has happened in the last week is a set of fascinating academic discussions about the difference between form and process, particularly with Lou Kauffman. 

The form of an experience is not the same as the process of experiencing it. The structure of the love scenes in the film is not the same as the experience of watching them. The form of a mathematical proof is not the same as the process of discovering it. The form of a piece of music is not the same as the process of hearing it, and certainly not the process of writing it. 

But form and process may be united at a deeper level. To think this is to think that there is no time. There is no moment where Maria and Tony are apart, and another moment when they are together. There is no "moment" at all, but everything is one together. To be in a deep spiritual state where we appreciate that is to have hope not just that things can get better, and that the moment of things being good and whole is ever present with the darkest of times. 

Wave Genetics, maths and Health

As is usual for this time of year, it's been hard to get motivated. I've got to the end of the week thinking that I haven't done anything, although one or two things I have done have been quite good. 

Involving Peter Rowlands in my academic department at Manchester I hope will be the most significant thing from this week. Peter and Sidney came to our departmental meeting - so what can a genius theoretical physicist contribute to our thinking about occupational health or public health? Well, one thing came up almost immediately - there is a deep interest in the University in the potential occupational hazards, and potential opportunities for improving OH in the widespread use of graphene. Any new material presents new kinds of risks, but to understand it, we need to understand what the material is.

In his book "Foundations of Physical Law", Peter made a prediction relating to graphene and the measurement of fermion velocity. Basically, one of the properties of graphene is that electrons (fermions) have no mass, and therefore can travel at speeds very close to the speed of light. Graphene presents opportunities for investigating the behaviour of electrons and quantum mechanics. One of the interesting empirical phenomena is the Quantum Hall effect, whereby the resistence of current flow can be seen to be quantized (i.e. goes in steps) in what are called Landau Levels as a magnetic field is increased. Peter's prediction is that as the magnetic field is increased, so the speed of electrons is reduced because they acquire mass. 

Might this fundamental property of the material relate in some way to the biological effects that the material might present? Or even, might the insight gained into the dynamics of quantised energy through graphene cause us to rethink the simplistic cause and effect models in occupational health or public health? I think both of these questions are very interesting. It all goes to show the power of an interdisciplinary approach to these things - particularly if we are lucky enough to get a concentration of highly original thinkers.

The second thing that happened this week was that I managed to finish a short article in response to a piece by Lou Kauffman on "naming" for the journal Constructivist Foundations. This week, Lou has also been leading a discussion on "Biologics" for the Foundations of Information Science mailing list, and gave a webinar which was attended by Peter Rowlands, Stuart Kauffman, John Torday, Plamen Simeonov, Karl Javorszky, Gordana Dodig Crnkovic and many others from FIS. There were one or two important things that came out of this which were new to me. 

Lou began with self-reference (as he did with the paper he wrote about naming), and quickly moved on to the difference between form and process. In the flow of conversation, Stu Kauffman introduced the recent idea of "constraint closure" developed by Mael Montevil and Matteo Mosseo (see Biological organisation as closure of constraints - Maël Montévil). I wish Loet Leydesdorff was still around - I think he would have been very interested in this. It also raises fundamental questions about the "self" to which a "self" refers in self-reference (which was one of the questions I raised in my response to Lou).

Then Plamen Simeonov raised the work of a Russian biologist called Peter Garyaev on what he called "wave genetics". Garyaev has an unusual take on genetics which took him to postulate mechanisms for genetic transmission far beyond conventional biology, some of which involved sound and music. This is obviously interesting to me - but it could be mad. But aren't the most interesting things like this? Wave genetics - Institute Of Lingvistiko-Wave Genetics. There may be a connection to Rudolph Steiner here - but why not!

There was a fascinating moment when John Torday tried to relate his thinking about symbiogenesis to Lou's thinking about the place of mathematics in biological systems. John's view is that maths is ontological in some way, and which is endogenised by biological systems through symbiogenesis, which provides the foundation for the cultural practices of mathematics that we know. This made Lou think and I suggested whether this idea of symbiogenesis and the embrace of evolutionary history as a dimension to the "constraint closure" might provide us with a way of uniting "form" with "process". 

A final question I asked related to what a protein perceives in another protein. Lou responded by illustrating a knot in a plastic chain (some kind of lego thing - very interesting), and saying that one could determine the topology of a chain by pulling parts of it. I think that's very interesting because the "floppyness" of the chain becomes a transmitter of information about the chain - so the inherent "disorder" of many proteins may present an invitation to be "pulled" by other proteins. 

Related to this was a comment made by Peter Rowlands about measurement earlier in the week: when we measure gravity, we are not measuring gravity but inertia. The same might be said for pulling on a protein.