towards a typology of people who feel really quite strongly about AI
the original reason i got into STS was because i was pretty obsessed with the transhumanist community. the philosophy of transhumanism itself - that it is morally right to use technology to alter and improve the human experience - is internet rabbit hole material, for sure. but the communities that coalesce around it are a whole other fascination.
in nov '22 - a couple of months into my social research methods MSc - chatgpt dropped, and my attention turned to the ideologies and communities surrounding genAI. so when Timnit Gebru et al started talking about the TESCREAL lot, i was thrilled that someone else had noticed the connections and similarities between the two groups (not to mention articulated them much more eloquently than i could!)
recently, i've been thinking about the various factions of ideologies concerned with AI, and wondering if anyone has made any attempts to map these camps. the closest thing i've found so far is a substack called a guide to the AI tribes - a helpful starting point, but very EA-AI-guy-in-SF coded.
i wondered if anyone else has tried to do a similar exercise, perhaps from a different geographic perspective. maybe this is something i can do some research into another time. for now, here are a few groups with strong opinions on AI i've seen out in the wild:
- EA types (i did the 80,000 hours course. i am not convinced)
- artists and creative workers
- "AI will take our jobs" in trad leftist/labour kind of way
- "AI will take our jobs" in a only-reads-about-AI-in-the-news kind of way
- AI ethicists
- AI nationalists
- AI safety type a: "let's engineer safer AI"
- AI safety type b: "let's govern safer AI"
- AI safety type c: "AGI turbo death is imminent"
- responsible AI types
- people who dislike AI but hate being called luddites and point to how much they love other kinds of tech
- people who happily reclaim the term luddite
- AI critical zine makers
- open source AI types
- sees AI as not artificial nor intelligent
- sees AI as alive and also a best friend
- sees AI as colonial infrastructure
- doesn't care, it just writes their emails
this is just a brain dump and not an earnest attempt at theorising. but i do think it's interesting how these varied conceptualisations of AI often cut across traditional political divides. it might be tempting to try and place people on a continuum from "pro-AI' to "anti-AI", but the reality is much more nuanced than that (the sociologist thinks things might be nuanced? oh i'm sure).
something i did like from the article i mentioned were the use of these three axes:
- AI is a sham <-> AI may soon become very powerful
- business-as-usual AI development is net bad <-> business-as-usual AI development is net good
- generally anti-tech <-> generally optimistic about tech
it would be helpful, in any future research, to consider the role of values when mapping these groups. what reason might a group have to tout themselves as AGI believers, to endorse AI development as a social good? for what reasons might a group generally be anti-tech? we are beginning to see a great deal of thought being given to the attitudes people have towards AI, but not so much to the personally and culturally held values that drive them. therein may lie the key to where allegiances might (could?) shift.