I turned again to doing liquid simulations this week. I’m convinced there must be a way to do it which isn’t stupidly, achingly slow and heavy and I was wrong. I asked on Instagram for some inspiration for future renders and got some really good ideas to help me so that gives me a week or two of directions to go but I’m going to stay away from heavy physics sims for a little while because they just take up so much time and space.
One of the best things I’ve read about AI is Is chess the drosophila of artificial intelligence? A social history of an algorithm by Nathan Ensmerger. The paper traces the love affair of AI researchers and chess and unpacks and questions the nature of this relationship; how it exposes biases about what computer scientists and mathematicians think intelligence, how it creates publicity and hype for the public and funders and how it provides some form of flawed benchmarking. It draws the comparison with genetic science and drosophila – fruit flies – and how their proclivity for morphogenesis, quick breeding and relatively long lifespans make them useful to the biological and genetic sciences. However, it disassembles this simile with the brilliant quote from John McCarthy”
Computer chess has developed much as genetics might have if the geneticists had concentrated their efforts starting in 1910 on breeding racing Drosophila. We would have some science, but mainly we would have very fast fruit flies.
John McCarthy, AI as Sport, 1997
In other words, the focus on chess as a benchmark of computation and intelligence has limited the possibility of AI to do anything else except conform to a certain competitive and elite understanding of what intelligence is or could be. So I really like this playful call from Sally Davies for Dungeons & Dragons to replace chess as a benchmark test for AI. She describes how the aim of D&D isn’t to ‘win’ in the highly constrained and logical ways of Chess and Go but to make creative stories, D&D players enjoy defeat as much as victory as long as it is supported by a great story. Then there’s the complex social entanglements of it; players know each other and are pretending to be others, imagining the actions of imaginative characters and intentionally trying to make it as enjoyable as possible for the group. It’s a good assumption that an AI that could be a good D&D player would be a lot better than one that can solve the mathematics of Chess or Go.
It also strikes me that it so easily could have gone the other way. In Utopia of Rules, Graeber writes that a very good explanation for the prevalence of magic and occult language in computation (setup wizards, mailer daemons etc. etc.) is because people working in computation were all playing D&D in their spare time. A computer is a mathematical system with a thin veneer of lore, after all.
Non-Aggressive Critique
I finally read this great interview with Clara Balaguer here with Rhys Atkinson, who used to be on MA Graphic Media Design where she talks about trolling design with ‘bad’ software and the fashion for exhausting words before they’re done but I really, really, really gelled with this sentiment;
Humour and self-deprecation is a non-aggressive approach to criticality… I don’t always use this tactic, but I can draw from it. And I do not see it as submission. It’s exerting negative power. By ceding with gentleness rather than occupying with aggression—with a joke or an apology or a compliment—one can command a space.
Which, you know is a big part of the Revell™️ brand. I’ve always been uncomfortable around aggressive critique, either people bullying others in crits or on Twitter dot com or this notion that critique is competitive and you are ‘more right’ by having read more and using the right words in the right order. I find satire and self-satire much more useful because it’s not simply about intimidating people. I was talking to someone the other day who said (paraphrasing) ‘When you go to the doctor’s you don’t expect them to reel off a load of medical research citations in order for you to trust their diagnosis.’
Recents
Well, hopefully you saw but I had a great chat with Jarrett Fuller on Scratching the Surface a few weeks back and that’s now been released. It’s here on Soundcloud but is available on all good podcasting platforms. We talk about university, design, administration, speculation and everything in between.
I need to do some site updates this week. There’s a list of recent talks, events, publications etc. that need to go up but I just need to do that yearly tidy and get them up there.
Oh, we’re looking for a new course leader of BA User Experience Design. It’s been weirdly difficult to find someone for this role but if you think it might be you or someone you know, let me know. You just need to know about UX, teaching and a little about managing.
Short Stuff
- Interesting piece on ‘Academia’s Stockholm Syndrome‘ and how we can’t manage to see past rankings. The author suggests that a) rankings are only respected and thus of commercial value to the ranker if they conform to what people already think which means the metrics used are chosen to do so eg: ‘Oxford and Harvard must be the best → by what metric are they the best? → Let’s use that to rank all universities.’ But b) that we can’t seem to see a world without rankings, griping and grumbling about the way they’e done as opposed to just rejecting them.
- The environmental cost of all this time on Zoom.
- Bruce Sterling has a new ‘art’ blog; Artmaker. Lots of stuff on NFTs. There’s not an easy way to subscribe or get those things in your inbox which is fine because I have 80 odd unread newsletters anyway.
- Firsting in Research from Max Liboirin about the problem of the reverence with which being ‘first’ to do/discover/see something is held in research.
Ok, that’s it back to work. I managed to read some stuff there but it’s late in the day now. I don’t call you enough. You know almost everyday I think ‘oh I gotta catch up with so and so and see how they’re doing’ and it’s very often you but events overtake me all the time. Love is a two-sided coin like that. Ok, until next time.