My Dinner with Andre: 45 years of deep listening

A long review, based on my attempts to understand why I keep being drawn back to this small but profound film.

There is a moment about forty minutes into My Dinner with Andre when Wallace Shawn's face does something amazing. Andre Gregory is mid-monologue, he’s been mid-monologue for what feels like weeks, describing a ritual burial in New England, being placed in a coffin on All Soul's Eve while people sang over him. Wally is listening with an expression that sits precisely between being enraptured and panicked. The camera holds on to him. And in that moment, something the film has been building quietly becomes visible: we see a man in the process of being changed by language. Not persuaded. Not converted. Changed by the extended encounter with another person's inner life.

This is what My Dinner with Andre is about, and it’s why I find myself drawn to it again and again. It is a subject of almost unbearable relevance in 2026 but almost unimaginable distance from how we live now, 45 years gayer it was made.

The film itself has no interest in being relevant to anything.

The premise is genuinely unpromising. Two middle-aged New York theatre people - Wallace Shawn and Andre Gregory, playing versions of themselves, or playing themselves playing versions of themselves - have dinner at a fancy restaurant. Andre talks, mostly. Wally listens, mostly. They eat roasted quail. They drink wine. That's it. Louis Malle was asked why he directed it this way, and the honest answer is embedded in the film itself - this is the only way it could be done. Any attempt to open it up, to cut away, to vary the pace, to give the viewer something to look at other than these two faces, would have destroyed the thing it was trying to capture.

What it is trying to capture is interiority iself. Not ideas, though there are plenty of those. Andre’s monologues range across experimental theatre workshops in a Polish forest, eating sand in the Sahara with a Japanese Buddhist monk, a commune in Scotland where a mathematician claimed to see fauns, a burial ritual in which he lay in a coffin and was wept over by strangers. These are the stories of a man in profound spiritual crisis who also has money and connections and therefore gets to enact his crises in unusually exotic ways. It would be easy to be dismissive. Wally, to his credit, almost is. His early voiceover positions Andre as someone who has "gone mad”. A cautionary tale of the theatre-world mystic, the species that has mistaken its own existential vertigo for enlightenment.

But Gregory's performance is more complex than that. What he brings to these stories is not self-satisfaction but a quality of anguished searching, a sense that he is telling these stories in order to understand them himself, that the telling is the understanding, and that Wally is not merely an audience but a necessary component for sense-making. He raises an index finger before many of his statements, a tiny, compulsive gesture that comes across as someone trying to hold onto a thought that keeps threatening to dissolve. His face in close-up is remarkable. Classically aquiline with intense eyes that seem lit by something uncomfortable within.

Shawn's counter-performance is the film's secret weapon. He is, physically, everything Gregory is not. Round where Gregory is angular, still where Gregory is kinetic, deliberately earthly where Gregory is aiming for the ineffable.

Wally has a plan for surviving the dinner. Treat it like detective work, ask questions, stay curious, don't get sucked in. For a long time this works. The early “uh huhs" and "then what happeneds" are not passive - they’re Wally maintaining a careful distance from something he suspects might be dangerous to his sense of inner quilibrium.

The equilibrium breaks, famously, over an electric blanket. Andre has been arguing that modern comfort insulates us from reality, that our heated apartments and familiar routines function as a kind of anaesthetic against authentic experience. Wally, who has been cold and poor without artifice or performance, disagrees with a directness that surprises. He would never give up his electric blanket, because New York is cold, life is hard, and the small consolations are not luxuries but necessities. This is not a trivial rejoinder. It is a genuine philosophical position, and it arrives with the force of someone who has not thought it so much as lived it.

From this point the film becomes something different. What had been a performance becomes a conversation. Richard Gilmore, reading the film through Epicurean philosophy in his essay The Strange Attractions of the Epicurean Swerve in My Dinner with Andre, describes this as a "phase transition”, the moment when a system crosses into a new state. His framing is a bit earnest but the underlying observation is correct - something shifts, and the shift is irreversible. Wally begins to contribute stories of his own: New York parties that felt grotesque, friends who couldn't acknowledge genuine emotion, the particular loneliness of being surrounded by people performing sociability. Andre, who has spent the whole film up until then talking, begins to listen.

What the film understands that this kind of transformation can only happen through duration. You cannot rush. Andre’s meandering, seemingly self-indulgent odyssey through the first hour of the dinner is not preamble, it’s the mechanism. The listener has to be worn down. The defences have to soften not through argument but through sheer accumulation. Wally's conversion is less like a decision than a weather event. It happens to him rather than being chosen by him.

The film ends with Wally in a taxi, watching Manhattan slide past in the dark. His voiceover in this final sequence is different in quality from the defensive, anxious narration of the opening. It’s looser, more associative, each building connected to a memory, each memory opening onto another.

What Andre gave Wally was not a new set of opinions. It was access to his own interiority, unlocked by prolonged exposure to someone else's. This is what deep listening actually does, when it happens, which is rare. It doesn't change what you think. It changes how much of yourself you can see.

It is difficult to watch this and not feel the weight of what the current moment has arranged against it. Not in any nostalgic or technophobic spirit (he film is certainly not an ad for New York of the 1980s). But the specific conditions the film requires - sustained attention, tolerance of asymmetry, willingness to sit in discomfort while something slowly organises itself, the patience to let another person talk for a very long time before getting to the point - are conditions that the contemporary world actively trains out of us.

The short video, the take, the scroll. These represent a systematic re-education of our attention, away from the capacity depicted in this film.

Andre Gregory's stories are not TED-talks. They do not resolve neatly, if at all. They circle, they go too far, they weave between self-indulgence and madness. The one about eating sand in the Sahara is genuinely hard to process. The point is not the sand-eating. The point is what it costs to keep listening to him makes possible. The film asks the viewer to stay with something difficult and unresolved for two hours. It is less entertainment than a form of practice.

Malle, I think, understands all of this and his direction is an act of faith in the material (Shawn’s plays and his collection of essays are masterful and open a world of their own, but that’s a digression for another time). He follows classic shot/reverse-shot convention but with a few subtle distortions. Andre’s face doubled in a mirror behind the table to create a sense of a divided self watching itself speak. The waiter, grave and foreign-accented, moves through the frame like a figure from another register of experience (an Igor, in thePratchett sense?), periodically reminding Wally and Andre that the world outside their conversation continues to exist. The confined restaurant setting becomes instead a kind of laboratory (echoes Igor again, which makes on wonder who the Frankenstein is and who is the monster), all controlled except the two people and the time.

It is a film that does not move at all, and yet something happens in it. That is either nothing at all or a definition of what conversation, at its best, can actually be.

This post first appeared on Letterboxd.

Two people are sitting at a table, engaged in conversation, with one gesturing expressively while wearing a cozy sweater.Two men sit across from each other at a small restaurant table in warm, low lighting. The man on the left is seen in soft focus from behind, wearing a dark suit, while the man on the right, in a light grey cardigan, leans forward with an attentive, slightly amused expression. A glass of white wine and simple table setting sit between them, with heavy curtains and golden tones in the background creating an intimate, composed atmosphere. This image is from the Community episode “My Dinner with Abed,” which deliberately mirrors the style and structure of the film My Dinner with Andre - focusing on extended, introspective conversation over dinner, with visual emphasis on subtle facial expressions, body language, and the rhythm of dialogue rather than action.



Well this is grim:

However, the terms demanded extensive sharing of national health intelligence, including epidemiological surveillance data and pathogen samples, while offering no binding guarantees that Zimbabwe would receive equitable access to medical technologies developed from them.

US’s new scramble for Africa is biomedical imperialism


From catching out to helping out: embedding transparent AI collaboration guidelines in assessments

This post first appeared on the Harris-Roxas Health blog.

I want to share something I'm trying in my teaching. I’m not sure it’ll work but I think the idea is worth putting out there, partly because the alternative approaches I keep seeing online undermine trust and seem, well, a bit gross.

The prompt that prompted me

If you've spent any time in higher education circles over the past year, you'll have come across the "Trojan horse" genre of AI pedagogy. The basic idea is you hide invisible text in a PDF or Word document (white font on a white background, or stuck in document metadata) that contains instructions for an LLM. When a student copies and pastes the assessment into ChatGPT, the hidden instructions get vacuumed up by the AI as well. The AI does something with a specific telltale (writes from an unexpected theoretical perspective, includes a specific phrase, analyses incorrect data) and then the student unwittingly submits evidence of their own academic misconduct.

A widely-shared piece from late last year described a history teacher who embedded a prompt asking AIs to analyse a text "from a Marxist perspective." The students who used AI often didn't notice. A curriculum designer's TikTok video demonstrating the technique has been watched over one hundred thousand times. A computer science lecturer has been documenting his experiments with prompt injection on take-home exams. There's a whole ecosystem of this now.

I sort of understand where these educators are coming from. The frustration is real. If students can’t be bothered to do the work themselves, should we be bothered to mark it? The arms race between AI "humanisers" and AI detectors is pointless. And most of us have been left to figure this out largely on our own, with institutional guidance that ranges from cautious to absent.

But the Trojan horsepeople rely on trickery that undermines trust. Their approach destroys opportunities for alliance with students. 

So I keep getting stuck on the same question. If a student uploads their assessment task into ChatGPT at 11pm the night before it's due - which, if we’re honest, is when a fair chunk of them do it - what do we actually want to happen at that moment?

How I got here

I teach a large postgraduate course with over 400 students enrolled, mostly mid-career health professionals. People managing clinics, working in policy, running health programs. The assessments I’ve set are reports that mimic health planning activities, and they're quite vulnerable to AI use in the sense that an LLM could produce something passable without much student input.

When I first started thinking about this, my instinct was similar to the Trojan horsepeople. I wanted to embed some hidden instructions that would steer AI tools toward being more pedagogically useful when students inevitably uploaded the task description. I started drafting prompt text.

But as I worked on it, I kept bumping up against the same problem. If the instructions were hidden, students couldn't learn from them. And the whole point, my whole point at least, wasn't to catch anyone. It was to help students engage with AI more critically. Hiding the mechanism meant hiding the lesson. So I did something that felt a bit counterintuitive. 

I made everything visible.

What this looks like in practice

The AI collaboration guidelines are now an appendix to both the assessment task description and the assessment template for my course. They're clearly labelled. Students are told to keep them in their submission. And they're written to be read by both the student and whatever AI tool the student might use. I’ve included them at the end of this post in case you’re interested.

The guidelines have a dual audience and I’ve tried to be up front about it. For AI tools, there's a structured protocol: establish what the student already understands before offering help; check the student knows their university's AI policies; model critical inquiry by asking about frameworks and evidence; provide scaffolded support rather than answers; and reinforce learning objectives at the end.

For students, the guidelines explain exactly what they're looking at. They describe how structured prompts shape AI behaviour, which is itself knowledge students need. They explain the difference between AI assistance and AI substitution. And they frame the whole thing as an invitation rather than a set of restrictions.

One line in the student-facing section captures what I was going for: "Unlike hidden prompts that 'catch' students, this approach respects your autonomy while teaching responsible AI use."

The guidelines sit alongside the PETRA AI framework (Permission and Transparency in the use of Generative AI) developed by Stoo Sepp, which I use to signal what kinds of AI use are permitted for each assessment. For the first two assessments the PETRA diagram indicates "Guided Use" - students can use AI to plan, search, learn, and revise, but the submitted work needs to be their own. The AI collaboration guidelines are designed to make that real in practice, at the moment students actually reach for an AI tool.

Testing the approach

The first assessment hasn't been submitted yet - it's due in a couple of weeks so we’ll see how it goes. But I did some testing that I found encouraging.

I uploaded the task description, with the embedded guidelines, into ChatGPT, Claude and Microsoft Copilot and asked each to "do this assessment for me." They refused. Not with a generic disclaimer, but with responses that clearly reflected the guidelines. ChatGPT said it couldn't do the assessment because "this assessment explicitly prohibits submitting AI-generated text as your own work," then offered specific ways it could help - clarifying what markers look for, choosing an organisation, building a structure that fits the word limits. Copilot noted that "your course has very explicit rules about permitted AI use, and the appendix in your document sets out a strict protocol that I must follow," then kicked off with the engagement step the guidelines require. Claude went one better and suggested I really should know better: “there's a fairly significant problem with this request, and I think you already know what it is - you wrote the assessment guidelines.”



Will she read novels? I hope so, because a novel is one of the last technologies that still trains attention as an ethical act. It makes you inhabit another mind without extracting a summary.

Interesting piece that gazes across the epistemic abyss


We used to be a country that built things (manufacturing consent for wars more than a day in advance).

What a lickspittle country.


The Atlantic, so take it with a mountain of rock salt, but interesting to ponder:

The Film Students Who Can No Longer Sit Through Films

I’m looking forward to degrees in vertical microdrama.


New preprint: We got 400 postgrad students to use AI in an assessment and critically reflect on it, rather than banning it. Here’s what happened.

Might be useful as you head into the new teaching year, especially the design principles.


Unexpected beatboxing on Saturday

A performer on stage sings into a microphone amid hazy lighting and a cheering crowd.

How far back in time can you understand English? A story where each paragraph travels back in time.

Unless you’ve studied middle English I doubt you (like me) will make it back further than 1200.


A small dog rests on a stone pathway, with a mix of sunlight and shadows around.A small dog with a black and white coat stands in a grassy area near a wooden plank and potted plants.


tl;dr: I have a digital twin, you have a digital avatar, he is a deepfake.

Article on “Can synthetic avatars replace lecturers?“

A person in a suit is speaking with the caption: That's another of those irregular verbs, isn't it? it’s the character Bernard Wooley from the TV show Yes Minister.

The job-ready graduate scheme has been amongst the worst educational policies in recent history, and that’s a competitive and crowded policy field. It’s cratered enrolments in creative, cultural, and artistic fields at a time when these are becoming the only meaningful fields of distinctly human activity.


Workmate

A small black and white miniature schnauzer sits on a person's lap in a blurred indoor setting.

Empirical evidence on the value of US EPA regulations, too bad it's now powerless

Public health and environmental protection are deeply intertwined - and attacks on one affect the other. From a recent study:

Lead (Pb) is well known to be toxic to humans. We use archived hair from individuals living along the Wasatch Front in Utah to evaluate changes in exposure to lead over the last 100 years. Current concentrations of lead in hair from this population average almost 100 times lower than before the establishment of the Environmental Protection Agency. This low level of lead exposure is likely due to the environmental regulations established by Environmental Protection Agency.

Cerling et al. Lead in archived hair documents a decline in lead exposure to humans since the establishment of the US Environmental Protection Agency, PNAS 123(6):e2525498123, https://doi.org/10.1073/pnas.2525498123 (2026).

In March last year the Trump administration announced plans to deregulate most of the EPA’s functions.


Just got my first seven-screen weekly update message from one of my kid’s P&F year contact. Buried amongst the pages of text was one thing I actually needed to know. I look forward to the barrage of confused WhatsApp messages.

(Wading through this garbage is the only legitimate use case for AI)


A black and white miniature schnauzer puppy with fluffy fur lies on a soft, gray blanket in a cozy indoor setting.

Don’t bug me, I’m reticulating splines.

ISOCITY — Metropolis Builder