Hello. In 3 hours at 3pm ET, I'd like to check in and chat with everyone especially about the introduction and chapter 1 of The Eye of the Master by Matteo Pasquinelli, but also broadly I'd like to hear what you're all feeling/thinking, and looking forward to or hoping to get to in this book.

(If you haven't done the readings yet, and want to have hot takes at 3pm about which chapters or topics you're most or least excited to read about, the introduction has (IMHO) a really clear roadmap of how Parts I and II are shaped and what they're going to be about.)

I'm super excited about this book after reading the introduction; the "micro-decisions" of driving that he describes resonates really strongly with even very recent critiques of how LLMs may do an adequate job of generating code on a per-function or even per-file level, but absolutely fail at the cognitive step that needs to happen to architect software (the thinking and planning that people learn to do, sometimes informally, whenever they start working on a project and think about how things should be structured - a critique that also applies to AI-generated essay-shaped things).

I feel like Pasquinelli arrives at a similar conclusion about how to draw a boundary around AI as I came to (regrettably unaware of this book at the time) when I was struggling to draw a distinction between state and administrative power vs "AI" either as we know it today or as far back as the 50s, or even going further back, as he does. I think I'm willing to stop at the 1950s in a way that neither Pasquinelli nor Dekeyser nor Blix & Glimmer were satisfied with, so it's interesting to push myself to broaden my framework for "defining AI".

I'm really really intrigued by the framework he offers in chapter 1, where he thinks about social algorithms, formal algorithms, and automated algorithms as three distinct things. Whether you agree with how he places examples or not, or even whether you agree with this framework itself or not, I'm at least really intrigued by this set of dividing points, and I see how these divisions elicit some "analytical texture", to use an idea that's been knocking around in my head lately, that maybe was hard to perceive before. I'd really like to hear if people have thoughts or examples that jump out at them based on framing algorithms in this way.

One last thought I'm going to tack on to the end of this meandering reflection: I'm so eager to read more about Knuth's "mission ... to make [Computer Science] into a respectable academic discipline". I've genuinely tried to write about the "project" or "mission" to make intellectual projects (like the study of algorithmic harms, or conferences/communities like FAccT, or CHI), and acknowledging or owning the intentional (maybe even ideological) nature of the project that basically asserts something along the lines of "this is a field of study around which a boundary can be coherently drawn, and this space is worthy of people's effort and labor; and also... maybe it deserves funding, and spaces or venues for people to publish, and maybe academic or industry recognition & support".

I would argue that every conference, every university program, every self-described or outwardly-acknowledge discipline is kind of its own mission to assert that the organizing principle of that venue (or institution, or community) itself (and probably some larger intellectual project) is a legitimate and worthwhile subject to pursue, and reading about Knuth on this mission in the 1950s to make Computer Science into something respectable is a really interesting phenomenon. And like... maybe today we can talk about whether Computer Science has achieved that (maybe in different dimensions; is CS "respectable" in industry? politically? intellectually? I think Durán would have said that CS has ascended politically recently, and I think given the composition of the US economy we would all have to concede that when the United States is "5 companies in a trench coat" and they're all AI or GPU companies, it's at least arguably the case that CS has ascended industrially/economically. I'd really like to hear from people outside of the CS discipline but in academia (whether k-12 or "higher ed") what perceptions are about CS.)

When you're ready to join at 3p ET, you can click on this link or you can go here:

https://al2.in/ReadingGroupRoom

See you then.