Where do we go from here?
Welcome to another edition of AI, Law, and Otter Things! Today's newsletter is somewhat self-centred: I will mostly speak about what I have done in the past academic year while taking some time to reminisce about TV shows. Next week we return to the usual schedule of more serious writing, but I think this filler episode might still be interesting.
(As an aside, I am not a fan of spoiler alerts. I might write a bit about that in the future, but spoiler-sensitive readers should feel free to skip to the next paragraph whenever I mention a work of fiction.)
Almost 20 years ago, Buffy the Vampire Slayer fans were treated to something unusual at the time (though not unprecedented): a musical episode. Season 6's "Once More, With Feeling" uses original songs to tell a story that explores the aftermath of the main character's resurrection and presents key themes for the remainder of the season. After the main plot is resolved, the main cast sings a final song, asking how the characters will deal with the secrets revealed during the episode. On that same note, I want to ramble a bit about my scholarly plans for next year.
Before looking at what lies ahead, I think it is interesting to look at the past 12 months. The PhD programme seems to be going quite well: I have written a May Paper that discusses how lawyers can analyse the life cycle of AI systems, which lays some interesting foundations I plan to develop over the next few years. Perhaps more importantly, the pandemic-induced complications did not prevent me from learning a lot from my peers and the EUI professors, be it at structured spaces such as seminars or through informal chats and social network interactions.
I have not done too poorly on the publication front, either, as I managed to get some papers published, and others are forthcoming. In light of the current global context, I collaborated with Juliano Maranhão on a peer-reviewed article and a book chapter on the data protection implications of AI in healthcare. While I do not plan to deal with healthcare themes in the future, this collaboration has been an interesting experience, as it led me to deal with domain-specific issues and their impact on the application of general norms. Another domain in which I have worked is legal automation: Maria Dymitruk and I have just finished a book chapter on the interfaces between data protection law and the use of AI to fully or partially automate the work of judges and their assistants, and I also worked (with Juliano Maranhão and Juliana Abrusio) on a Portuguese-language review article of AI regulation and legal AI techniques.
These works on specific domains draw heavily — at least on my side — from my research on the interplay between technical and legal concepts in AI regulation. Since this topic is the core of my PhD thesis, I still have much work to do on it. Nevertheless, this general research has already contributed to some intermediate results. I presented a paper at ETHICOMP about the conceptual problems we face when describing AI systems as "algorithms", which I plan to further develop into something publishable. I also worked with Giovanni Sartor and Juliano Maranhão on contributions to a forthcoming GDPR commentary: data protection by design and by default (Article 25), pseudonymisation (Article 4(5)), and the legal bases (Article 6) for content personalisation in online platforms. All in all, it has been quite a mind-opening year.
I mention all that not just because I want to brag (not that I am above it, of course) or because some readers might be interested in one paper or another. Instead, my reaction when I had to prepare my yearly report to Fundación Carolina was to think, "OK, what can I do better next time?" And that has helped me set some goals and expectations for the 2021–2022 academic year. Perhaps the most important of them is that my rhythm in the past year is not exactly sustainable. While my research at the EUI has helped me mature my thinking and improve my outputs, much of what I did in the past year was a (substantial) development of lines of research that I started exploring years ago. But my research for the remainder of the PhD will take me to new paths, and it will be essential to leave some time to become acquainted with new methods, new theoretical frameworks, and even to explore some trails that might not be fruitful. As Tony Hoare put it, "premature optimisation is the root of all evil."
I will also need to take more time to adjust to the dynamics of legal academia. One of my priorities towards this goal is learning how to make myself heard, that is, finding the venues and the approaches that increase my chances of being listened to by fellow scholars. Of course, it is not always easy to see if a paper has had any impact, but my main interest in academia is in participating in the exchange of ideas about my research topics. Hopefully, I will get better with that as I mature as a scholar.
Another relevant cultural norm that I must attend to concerns authorship. As my overview shows, I thrive in collaborative environments, which allow me to test and mature ideas with my colleagues. However, single-authored papers seem to be quite common in legal scholarship, while they are becoming rarer in computer science and other disciplines such as economics. This means that I will need to balance my fruitful experiences with collaboration and solo work and what I am doing for the thesis. Welp, every choice has an opportunity cost, I guess.
Now that the rant is done, I must admit that I originally meant to write something about the latest Rebuild of Evangelion film. It would be a good thematic fit for this newsletter: not only Eva has some personal relevance for me, but one of the major themes of the movie is precisely about moving forward. But now I have written too much already, so I will just recommend the movie to anybody who has watched the anime.
See you next week!