Thank you to folks who sent me stuff after the last post. It was really nice toget some reading and thoughts and suggestions and a weirdly interesting response on LinkedIn(!) which is not normally where the interesting conversation is but I guess it’s all the same people.
Making that big map of stuff was incredibly helpful. I mentioned last week that one of the problems I’ve had is that in my head, all of these ideas are cyclical feedback loops; ‘AI as-is shapes our imaginary which we act on through pop culture which influences AI as-is which shapes pop culture, etc. etc.’ And that’s just one example. None of this stuff is linear and so I was really daunted by how to take really circular ideas and arguments and write them in a linear way; with a start, a middle and an end.
Whatever it did, putting that map down helped me get to a point where I could trace out the loops in the ideas. I could then take what felt like the big ‘areas’ and draw them out into the above which become, effectively, chapters as you can see above. This helped me understand that they could still be feeding into each other but still approached sequentially, like a spiral; generally moving forward but with iteration over previous notions or ideas. Ok I just realised this is quite vague, what do I mean?
Well, take for example ‘Enchantment.’ Each of the chapters is a dimension or aspect of the way in which imaginaries of AI is constructed. Enchantment is one; the sense of this sublime, superhuman power that can do ‘weird’ things. Part of its enchantment is in its apparently incredible accuracy; that it can apparently make incredibly accurate predictions. But these predictions are in themselves, generally, highly-choreographed performances in which there are either strict constraints on the variables and these incredible results are deployed as a spectacle of marketing. These spectacles in of incredible prescience get so much attention in the field that they create an imperative for different actors, developers and companies to compete for even greater results, further distorting the benchmarks they are testing against in the pursuit of ever more accurate results, to increase the sense of enchantment, through the spectacle.
However, the remarkable success of these tests also tied up in the invention of users and use: So your AI is great at chess? Well chess isn’t exactly a multi-billion dollar market. So lessons learned or results proven from highly limited benchmarking settings are manufactured to have real-world applications and users and use-cases are invented; predictive policing, autonomous vehicles and facial recognition being the three top ones probably. There in turn set new benchmarks for the market as companies out-compete each other for ever more accurate systems and round and round it goes.
So these things feed into each other in a constant spiral but seeing it as more of a Venn map like above makes it easier to think about navigating my way through.
What happened to the pretty pictures?
Another big thing is the absence of CGI. For ages, in my head this whole thing was about AI and computer graphics but in actually going through stuff and really thinking about it, CGI is just a component piece of the design question which the thesis is centred on: The designerly element here is the way that these dimensions or aspects reinforce certain norms, assumptions or biases and how design might be used to challenge them or imagine alternatives. Some of that might mean CGI, some of it might not.
This of course means revisiting the abstract, which is fine because that’s what an iterative process is for and making that decision to cut off or at least demote the importance of something i really love talking about is fine because it’s not like I have to stop talking about it, we just need to make some decisions.
Another thing lurking at the edges is the sort of AI ontology stuff. Now, some of that is in there, particularly a lovely little thread about how prediction as a future notion shuts down imaginaries which I love: ‘There’s no point imagining alternative futures because soon prediction will be so good that we’ll know the future’ is the very, very simple version but this puts it much better. But perhaps something running across everything is the type of knowledge and imaginaries that are produced by an AI-centric technoculture, not just the imagination, but the very operating logics. One to keep mulling on.
So the last few days have been iterating on the chapter headings and content descriptions which is surprisingly straightforward: A description of the dimension and aspect and why it’s important and then the implication or role for creative critical / designerly practice. Hoping to have that ticked off by end of week with some other introductory content on definitions (what is constructivism, what is AI, what is worth talking about, what isn’t) and then I can finally make Stopher happy by reading his literature review.
Upcoming
Abundance is going to be shown at the Design for Planet festival 8-9 November. It’s a great lineup of people anyway so you should try and attend.
Short Stuff
I wish I had some but I’ve been lax on reading or doing anything extra curricular that’s not related to the PhD work. I could always just list the things I’m reading there I suppose. Ok, love you, speak later. I found out that that stalker guy who went to prison signed off his vlogs saying ‘I love you’ which has put that notion in an uneasy position now in my mind. But Adam Buxton still uses it and that’s really the vibe I’m trying to emulate here.
I’m off to hospital again this afternoon to have my leg scanned and be told everything’s fine over a four hour period so that’s a good chunk of time to hopefully get closer to finishing Dawn of Everything. Ok. I do love you and I will write you next week.