Agora U

The Provost, Fellows, Foundation Scholars and the other members of Board of The College of the Holy and Undivided Trinity of Queen Elizabeth near Dublin aka TCD, like other Institutions across the world, is should be having an existential crisis in response to "AI". But Institutions have the inertia of a super-tanker: they keep going in the same direction because change is just . so . hard for the [old embedded complacent] people who hold {power | purse | policy }in their hands. In my section of social media [<coff> Metafilter] you have to be super-tippitoes if you want to advocate positively for AI. At least partly, this is because "AI" is not one thing, but a range of tools created by a range of actors, only some of which are The Axis of Evil. But what do I know? AI has tsunamied over me (and you) so quickly, so recently, that I am reduced to slogans with as much discrimination [both senses] as "Four legs good, two legs bad".

In 2018, the Fellows of TCD elected immunologist Cliona O'Farrelly [my old boss] as first female Chair of Fellows. In 2021, they voted for Linda Doyle, an engineer, as the first female Provost in 400+ years. On 22Apr26, as part of her exit strategy, the outgoing Chair of Fellows organised a symposium of Two households Six pundits, both all alike in dignity, In fair Verona Dublin, where we lay our scene on "AI and the Idea of a University". I booked in because it seemed like a good opportunity for Me to find out how to spell AI and learn what my Alma Mater was going to do about it: not least because the first speaker, after the Chair, was The Provost.  

Prizes:

• Most obscure cited reference John Kelleher for Laboratory Life: The Social Construction of Scientific Facts (1979) by  Bruno Latour and Steve Woolgar. 
• Also for best $10-word: semiotics - see Umberto Eco 
• Best timekeeper (by far) Camilla Persello, Secr of Scholars, also by far the youngest panelist 
• The Whitehead Prize for talking waaaay over the allotted time?  redacted .
• Fewest on-message bullet-points (3) The Provost:
• Are we good enough? - the antidote to smugness
• Dreadnought AI - it's here & not going away: engage challenge resist
• Truth is grey not black&white
• Most embedded in certainties (and not in a good way, although my bloboprev guts are with him generally) Fintan O'Toole
• Also for mentioning Tim "The are no Short-cuts" Robinson whom RIP
• Most empathic speaker Jennifer Edmond - wish I'd been a student of hers.
• Shortest on-message comment from the Floor Ken Mealy
• Most valiant effort (largely successful) to rein in his tendency to ramble: Ian Robertson [whom bloboprev]. 

Exec Summ [filtered through my jaundiced eye]: 
[I was far better at taking notes in 1976: in 2026 I found that an idea I was in the process of capturing on paper was slipping beyond my ken because the next idea was demanding my attention. Sorry if I've missed something important in the discourse. Doubtless within a few days I will be inventing whole paragraphs]  
In my day [1973-77], teaching at Trinity was very heavy on the Medieval model: The Professor / expert stands by a chalk-board and imparts his [almost always His] accumulated experience; know-nothing students write down the content with a pen, in a book. Weeks or months later, Prof sets and exam to see how well students have retained his imparted knowledge. Students with the most legible notes and most retentive memory are rewarded with high grades. The highest graded students become professors in their turn. Nobody taught us Pittman's short-hand

At the AI-fest, Camilla Persillo pointed out that group learning, where students bat ideas around together without faculty in the room, can be the most exciting and productive learning experiences during college days: it's the synergy innit. Several of the other contributors, incl Ken Mealy, made the point that Content is So Yesterday; nobody now needs to take notes to recall Scads of Stuff when they can look it up in two tics on their device. The trick is a) to remember how to question the source b) have a well-polished crap-detector to critically evaluate the 'answer'. If the curriculum and the exams require memorizing Avogadro's, Bernouilli's, Charles', Dalton's, Euler's, Faraday's . . . Laws then there is no time left to teach ascertainment bias, availability error, anchoring bias, authority deference and other cognitive fails.

But whoa! Only some 21stC students have the time for group study with peers. It's fine if you are a Foundation Scholar with free food and rooms in College, or if you live at home a short jog from the city centre because your family's generational wealth is a lovely red-brick in Rathmines. Not so much if your commute is 90+ minutes to Outer Boondocks and/or you have to slave in the local Spar convenience store several evenings a week to make rent and/or care for your beloved but demented Grandpa. Inequality [of resources and opportunity] is a systemic failing of our FF/FG society regardless of AI. But AI provides an option for the dispossessed to help them knock-off assignments which they have no time to address in a medieval collegiate manner. 

Another much worked point was the desirability of being uncomfortable with your data or ideas. The phrase used is meaningful friction . . . and b'god it involves Effort. Prof Robertson asserted that the cognitive work of [trad] learning increased myelination and connectivity of nerves in the same way as weight-training or running increased the # mitochondria in muscle cells. [Protestant?] work makes you fitter and cleverer. Using AI for cognitive-offloading, let alone cognitive-surrender is a disaster for your education: over-use of AI might make you stupider after college than before. Prof Edmond's experience is that the smarter students work effectively with AI to produce even better deliverables . . . but the weaker [tired, poor, huddled] kids turn in shoddy because they don't have the spoons to discriminate own-self-okay from superficially convincing AI-slop.

All agreed [everyone present being invested in The Idea of the University!] that Agora University was first-and-foremost the collective intellectual interactions of the people [provost, fellows, junior faculty, scholars, students, that janitor who nailed calculus problems] present on Campus. It was also recognised that (in addition to reviews, recordings, re-thinkings; patents, papers, plays; dissertations, discoveries and degrees) the deliverable is Citizens. The more engaged, ethical, thoughtful, kind, inclusive, resilient, the Better. 

[[Personal sidebar: I was entirely off-with-fairies during my last two years studenting at TCD. The most exciting thing that happened to Evolutionary Biology in the 1970s was Sociobiology and the genetic basis of altruism. I spent several days out in the library of The Other University because they had the academic journals in which this material was published. I read, and wrote a long-form review of, EO Wilson's 1975 book Sociobiology. I've written about my off-piste investigation of inbreeding and fertility in the Habsburgs. I was stoked! But I was also failed! when it came to the exams. With different teaching methods and more me-adjacent learning opportunities I coulda been a contender for the next professorial vacancy.]]