What the FAccT?

Hello hello everyone — a couple of weeks ago I was at FAccT, and I read loads of interesting research papers, lots of which had relevant findings on technology politics issues, and some that were too wonky to really absorb. After spending a week surrounded by smart people working on issues I’m interested in, I have some thoughts on the general state of this type of technology research.

If you prefer the medium of voice, I actually had a great conversation (at least I enjoyed it!) with Andrew Strait about our favourite papers from the conference — listen to the podcast here. Whether you have FOMO about it, or just want a few academic bon-bons for your amusement, check it out. There will also be three more episodes from my time at FAccT, where I interview some of the researchers I met — this Friday’s ep will be about abandoning algorithms.

🎙️ This is also just a reminder that you can get our podcast on Apple/Spotify/your preferred platform, and you can give it a 5/5 rating; you know… if you want.

And now, some thoughts on FAccT

The Fairness, Accountability, and Transparency conference is a little bit weird. It’s an academic inroad to a lot of social and political questions about technology. But it’s really hard to separate out society, politics, and technology, so producing papers from specific disciplines can flatten a lot of the complexity of the issues. But the disciplinary rigour isn’t always a bad thing.

FAccT research and conversations can focus on the technical aspects of fairness, accountability, and transparency in a way that is maddeningly narrow as to be almost irrelevant, and also in a way that is refreshingly specific and more likely to influence policy and product.

At the same time, applying a sociological lens to questions about technology can be dizzyingly expansive. It can call into question the very premise of technology and its underlying ideology in a way that transforms the depth of all possible political conversations. But that also means it can be so esoteric, that it can be almost impossible to act on.

Landing on actions or solutions is often Not A Thing with this kind of research, because researchers get bogged down in locating and analysing problems — and that was definitely the case here. But, even so, I found the problems that participants focused on to be so worth engaging with. Even when over-intellectualising what should be done about an issue, I still found it germane and helpful.

Simmering in the background was also the ideological battles between effective altruists — who conceptualise harm to humans as an abstract numbers game — and the activists and researchers fighting for the dignity of their fellow humans. The former focuses on problems caused by technology by using more quantitative and removed ways of knowing. The latter, on understanding the lived experiences of those impacted by technologies. It’s hard to imagine reconciling those views, but trapping them in a weird hotel in Rio for a week and inviting them to learn things from each other created interesting moments of friction and maybe also appreciation. I for one had probably my longest conversation with someone from the AI safety community, and found it really informative.

There were loads of other tensions too: industry-aligned and industry-challenging; policy wonks looking for solutions and academics looking for knowledge; product people and infrastructure evangelists; incrementalists seeking to help the powerful be more fair and radicals working out how best to wrestle power away from those who have it.

I wasn’t presenting a paper because I am not a researcher. But I wasn’t the only sicko there just to learn. Since I first attended FAccT in 2019, I feel much more diversity in the types of people that attend. I think this adds so much. This novelty and dynamism was contrasted pretty strongly with the dry, academic proceedings, which included calls to separate politics from science (lol), and was kind of disorganised, despite the amount of free labour from the amazing volunteers running the show.

Overall I had an amazing time, and I think FAccT honestly is unique in how (in such a large space) academics, activists, data scientists, and social scientists are thrown together to discuss and explore deep questions about how technology and society should — and shouldn’t — co-mingle.

Thank you for reading what was essentially a Yelp review of a tech conference — as usual, you can reply to this email and tell me what you think. Did you go to FAccT? What did you like about it? What did you HATE about it?

Don’t forget about our Insight Session on how the NYT reports on AI — happening tomorrow!

We have Hanna Barakat presenting her AI content analysis of the New York Times; she will take us through the prevalent themes and narratives present in the NYT’s reporting on AI, with a wider goal of expanding this analytical framework to other media outlets.

👉 The session is tomorrow, 26th of June at11am EST/4pm BST. Register here 👈

Hope to see you there!

Alix

If this was forwarded to you, sign up here.