Dougal McNeill

Designing for friction. Vivienne Ming in the FT:

“A Harvard study published in 2019 captured this in a single, counterintuitive finding. Students challenged to wrestle with problems learnt significantly more than those in traditional lectures, yet they reported feeling as if they had learnt less. Our brains mistake the smooth, fluent sensation of being told something — whether in a lecture hall or by AI — for the harder, messier process of actually learning. And generative AI is the most fluent thing that humans have ever built.

To resist that requires choosing productive discomfort: being wrong, or doing the unglamorous work of interrogating an answer that looks correct.”

“It would seem, therefore, that the real objective of the Crown’s proposals is to reduce the legislative Treaty protections available to Māori”: Carwyn Jones’s evidence to the Waitangi Tribunal on proposals that “will, if enacted, constitute the most wide-ranging legislative breach of Te Tiriti in modern history.”

If you have ‘40 000 words of your own original materials’ to hand you can always use them to, uh, write your own work. It’s not just that I don’t want CoPilot to write my emails or summarise documents for me or any of the other nonsense (I want it to leave me alone). If it’s your ‘extensive base of knowledge’ (what are these words!): you can summarise it yourself. That’s hard sometimes, sure; it’s also fun. And it’s how ideas happen.

Sydney academic used AI to write SMH opinion piece urging students to avoid using tech to ‘cut corners’ - The Guardian report on Western Sydney’s Pro Vice Chancellor for Quality and Integrity (I know) and her, subsequently withdrawn, CoPiloted _Sydney Morning Herald _ column.

There are basic, foundational, fundamental questions about quality and integrity raised in this article and in the responses quoted in it.

Anne Kennedy picks Bill Manhire’s ‘An Inspector Calls’ at NZ Poetry Shelf for ‘playing favourites’. She praises ‘the jazzy sound of it, the funny rhyme at the beginning, the noir feel, the angular look’. Yes!

Readings on Lygon Street in Carlton used to sell the London Review of Books, and I’d pick up a copy every second month or so and read it at the pub across the road, on the corner of Lygon and Elgin Street. What was it called? I can’t remember. That, I’m sure, is where I first read Manhire’s poem - in that pub, printed in that paper - and the last line stuck with me since then. A Peter Steele poem from the same year - published in The Age? - ended, I’m sure, with the line ‘but this is Carlton’. I’ve never been able to find it since, and don’t know if I’ve imagined it now.

The pub is a vegan restaurant now, the tatty and cavernous second-hand book store that was next door to it burned down. Readings, last time I went, don’t stock periodicals:

There’s always a point at which a routine enquiry

turns into something else entirely

‘Despite the braying of the tech elite, we still have agency. We still have a choice. Players with many billions at stake have a vested interest in removing your agency, and reclaiming it hurts their bottom line. There’s no way to say for certain who will be right about AI in the end, but the current evidence points towards disaster. And it’s safe to acknowledge it.’ Marisa Kabas with some good hating.

‘The speed of the digital world connives to make us feel ashamed of certain sorts of slowly won knowledge. But to persist in caring about books, and to do so in the face of those who tell us not to, is to fight for a world that takes knowledge seriously. That fight will have to happen on many fronts; preserving the library alone will not rescue reading. But it is a good place to begin.’ Sheila Liming on ghost libraries.

‘So, why save the university?’ Lorna Finlayson on the how the British university is dying, and the ‘colossal and entirely avoidable act of self-harm on the part of a country [Britain] whose higher education sector has until recently been one of the few things it had going for it’

More from Magnifica Humanitas:

‘We must, then, avoid the “Babel syndrome,” namely the idolatry of profit that sacrifices the weak, a uniformity that neutralizes differences, and the pretense that a single language — even a digital one — can translate everything, including the mystery of the person, into data and performance. The risk of dehumanization — of building a future that excludes God and reduces the other to a means — is an ancient and ever-new temptation that today takes on a technical guise. Instead, let us choose the “way of Nehemiah,” which highlights the importance of working together to make the City of God a safe place for returning exiles. Rebuilding today means recognizing that, precisely from the plurality of voices and visions which, even though they sometimes remind us of the confusion caused by the diversity of spoken languages, a bright possibility emerges. Indeed, this is the possibility of building together, of transforming diversity into a resource and of making listening and dialogue the common ground upon which to cultivate justice and fraternity. Within this shared task, Christians discover their unique role of guiding actions toward God so that, in his light, pluralism does not dissipate into disorder, but instead, through the practice of synodality, it becomes the space in which humanity rediscovers its solid foundations and its final end.’

Texts read us, and genres train and programme their readers. Some of the basic questions from literary studies playing out now across extra-literary human relations, and in ever more worrying ways. This from Myra Cheng and Dan Jurafsky’s work at Stanford:

“Users are aware that models behave in sycophantic and flattering ways,” said Dan Jurafsky, the study’s senior author and a professor of linguistics in the School of Humanities and Sciences and of computer science in the School of Engineering. “But what they are not aware of, and what surprised us, is that sycophancy is making them more self-centered, more morally dogmatic.”

Informed comment by Alexandra Sinclair, writing in The Spinoff:

‘There is no public mandate to cede huge amounts of power in designing government infrastructure and providing government services to overseas technology monopolies. New Zealand’s data sovereignty context is highly distinct. AI models licensed from overseas are not designed for an indigenous context. There are also significant data sovereignty concerns with training overseas general-purpose models with Māori data.’

Encouraging reporting by Dana Goldstein and Lauren Lancaster in the NEW YORK TIMES this weekend, and Philip Matthews' profile of Karen Hao, ‘The age of slop: the growing backlash against AI’ in THE POST, both have me cheerful. Goldstein and Lancaster describe moves by teachers in college and school classrooms in the States to do things together, as communities of learners, that help: reading, writing, thinking, engaging. Chat-GPT (or Claude, or CoPilot, or any of the other slop they’re pushing on universities, and that is in turn being pushed on students) doesn’t repel me because it’s an aid to cheating. It repels me because what it produces is boring. And ugly. And dull. And - most important of all! - a _barrier _to learning. Why facilitate the adoption of something that’ll make us all less free? Pens and paper and scribbled-out lines and drafts and false starts don’t represent a return to anything old-fashioned. They show a revival of commitment to the process of reading. That’s fun! I don’t want my computer to generate a bullet-point summary of a thesis I’m about to examine. I want to be left alone to read. And I don’t “deliver content”. I teach. And, in teaching, I learn.

“AI” is a big deal and important for all kinds of (alarming) economic, social, political and political-economic reasons, sure. Anne Alexander and Holly Lewis (and Karen Hao) are my guides for all of that. But, in the literature classroom, it’s something much less interesting: a scam and a distraction.

Christina Gratorp (who was a coder for fifteen years before she started a PhD in Environmental and Energy System Studies at Lund University), has written a brilliant chapter on ‘The Hidden Labour of Automation’ (published in the book (DE)AUTOMATING THE FUTURE: MARXIST PERSPECTIVES ON CAPITALISM AND TECHNOLOGY, 2025). She writes about everything from the development of the microwave (and how it didn’t reduce unpaid domestic labour) to the place of the messenger boy in the era of the telegram. The idea that, somehow, somewhen, the technology will sort out social problems is, she points out, an ideological myth, not a social policy programme: “the present is always in a debugging phase.” Nothing is really automated, just like there is no such thing as software. The work, the manual labour, is happening somewhere in the cycle, just out of sight.

“Even if some automated processes can save time,” she concludes, “focusing on this distracts from the proper goal of reconstituting technosocial relations. Treating everyone’s time as equally important - which in the proper sense of the word would be radical - instead harks back questions about how work should be organised and how it can be sustainably, equally, and fairly distributed.”

(These thoughts prompted, in part, by the Canvas shutdown.)