I’ve been increasingly seeing the idea of ‘speculative design for preferable futures’ appear in various places and thinking about why it sounds so discordant and icky. A big part of the critique emerging about the RCA version of speculative design over the last decade or so was the tendency to aestheticise dystopia and focus on technology – both of which are fair. However, the methodological approach (if there could even be said to be one) is very much about being open and exploratory, best represented by the rubric of ‘what if-ing’ that dominates that canon of work. There’s no particular intended outcome or impact (something I’ve also thought about before) other than asking people to ask questions of their assumptions about the future.
By steering the imagination towards a preferable outcome, do you provide an inauthentic critical space for the future and also prescribe in others a particular future vision? – The old ‘preferable for who?’ A lot of these ‘speculative design for preferable futures’ projects seem to be driven by participatory or co-design practices which I suppose does invite greater authorship and Pedro Farias has done a great job untangling some of those complex ideas of design, futuring and participation. My worry is that these preferable futures, though important, distract us from confronting undesirable futures or presents and white-wash the problems.
For instance, a project on ‘preferable futures in healthcare’* might suggest a way of getting to a fair and equitable future for the NHS (which, for the sake of this, let’s all agree is ‘preferable’) but in focussing on that future and the pathway to it, brushes over the political and economic conditions of the NHS as ‘solved’ for the sake of the preferable future: It fails to acknowledge or grapple with the politicisation of public funding for healthcare in the UK in the same way that a project that acknowledges an undesirable (and equally likely) future in which NHS privatisation and defunding continues, a la Dynamic Genetics vs Mann which I worked on with Superflux back in the day. This, in a way, is just poor design, because the politicisation of the NHS is one of the greatest ‘pain points’ (in service/UX design terms) of the problem. By making it about ‘preferable futures’ you keep the design process from engaging in the political reality of the problem; something that UX/service design is always really anxious about acknowledging.
Ironically, the focus on ‘preferable future’ does the same job as the ‘promissory rhetorics‘ I talk about so much in my PhD work; the use of the ideal future as a cover, blinder and distractor for the equally, if not more likely, un-ideal future: ‘Yes there’s a good chance it all goes horribly wrong and we end up with mass, centralised surveillance built on biased data sets, but what if it works?!’ The gambler’s fallacy of futurism. Ironic because I’m such a bravdo-driven optimist.
PhD
So the last two weeks have been on what could loosely be called a ‘literature review’ which comprises part of the context-setting for the thesis. The context-setting bit of the thesis is in three parts: First, a general review of the contemporary critical theory and practices around AI. Secondly, the theories I’m using to think about the way AI interacts with society and imagination (social construction of technology, critical futurism, epistemological cultures and others.) Finally, a justification and review of design practice as research and practices adjacent to what I’m looking at.
The first part of this context bit was what I was working on over the last two weeks. This involved going back over all the ‘AI’ tagged things in Zotero (about 400 items) and the brane to try and find a way to describe all of that stuff. I think I was always nervous about doing this before because it’s so impossible to even feel like you’re scratching the surface of the literature let alone being exhaustive so I just tried to signpost the main ideas and arguments going on and called it good. If someone wants to jab me about it, then fine.
I started with lots of small, single paragraph ideas like ‘Opacity as an artefact of data complexity’ or ‘Environmental impact of large data sets’ – probably about 20 or so in total. However I found I was looping back and forth between them too much as they all relate and so they ended up coalescing around three larger subsections:
Reductionism, prediction and determinism looks at the literature and work on the sort of ‘AI worldview’ – where everything is computable and reducible to simulation. This also encompasses the enchantment of prediction and the nature of the knowledge ‘fabricated’ by AI. Expectations, vision and reality looks at the work on hype, rhetorics, language, metaphors as well as opacity and intuitiveness to human understanding. Ethics, legibility and accountability looks at, well, the ethical debates around data, algorithmic accountability and so on.
What’s great is that for the first time it’s clear in my mind how these context parts (including the two yet to come) fit in with the body of the work, which is great! For every other draft of this thing, the approach has been ‘here’s some stuff I know. Not sure why. Not sure what it means.’ Now it’s; ‘I need to explain this to you so you understand why I’m talking about it in this way later on, ok?’
I spent the best part of the week assembling the pieces of this chapter and then on Saturday just full-on ran at it for about 12 hours and felt fucking terrible afterwards. It was incredibly frustrating; I went back and forth moving paragraphs and references around for hours and was going mad near the end as the time ticked down to baby-feeding time. On Sunday afternoon I reasoned it had to be finished whether I was happy or not and so I took another punt and was really pleasantly surprised to find that I actually liked most of what was there and that – on reflection – I’d done quite a lot.
Normally I nibble the way at the edge of the thesis; 60 minutes or so every morning. So the sense of gradual progress is really there and I feel good about every few hundred words added. It’s paradoxically much more rewarding (and probably better for quality) to break it up into small, manageable tasks I can be happy with. If at the end of a session I can say ‘Yeah, I’m real happy with that paragraph – it fits nice,’ that’s better than when I left it on Saturday night going ‘That 5000 words spaffed all over the show is a fucking catastrophe.’
So anyway, that’s done. Still room for improvement of course and I’m still reading so it will keep growing and probably morph a bit as new stuff comes up (again, the frustration of writing about something that changes every week.)
Recents
Busy month for Revell™️ actually. Kyle Goodrich’s book Computer Generated is now available for pre-order. I wrote the foreword about the near-history and future of CGI artwork and some of the critical dimensions of the technology. I was back on with the From Later crew and Radha to talk about #breezepunk. (I’ve put the video below.) And finally, the recording from mine and Natalie’s appearance with Kara Chin’s exhibition has been released. We talked about, well, yes, CGI and simualtion.
Short Stuff
- The DCODE summer showcase is up. Probably some of the most interesting critical research in AI there is. I need to look back over it actually for that literature review.
That’s it I’m afraid. I have a bunch of tabs open and really thought I’d have time to glance at some but I’ve had a tough week leg-wise and flu-wise and any spare time I did have was on the PhD or child-rearing so.
As I was holed up sick in bed for a bit this week, I was watching Justin Hawkins rides again. Justin Hawkins (of The Darkness fame) basically looks at, analyses and speculates on songs for ten minutes and was reminded to renew this idea of doing something similar and accessible for design. I don’t know if my vanity stretches that far and it’s not like i have time.
*Yes, this is a real project I read about, no I’m not going to name because it was actually a pretty good project and I just felt personally niggled by it and also it was a student project and students should feel safe to try things out and make a name for themselves without old cynic like me cussing them out to their friends.