Slow thoughts

My paternal grandfather was a captain in the merchant marine, as was his father, and his grandfather before him. In the room where I used to sleep, he had an embroidered motto hanging on the wall. It read:

“A true sailor doesn’t ask for the wind to change; he learns to set his sails.”

I always thought of that as vaguely masculine, bordering on macho — which is also how I thought of my grandfather. It never crossed my mind that the deeper meaning might be metaphorical.

Never, that is, until I stumbled on a passage in Samantha Harvey’s Orbital. In it, Nell — an astronaut in orbit — thinks of her sheep-farming husband below, and the contrast between their two lives opens into something larger:

Her husband says that Africa from space looks like a later Turner; those near-formless landscapes of thick impasto shot with light. He’d told her once that if he were ever to be where she is, he’d spend his whole time in tears, helpless in the face of the earth’s bare beauty. But that he’d never be where she is because he’s a man who disappoints himself with his need of firm ground. He needs stability inside and out, and to simplify his life lest it overwhelm him. There are people like him (so he says) who complicate their inner lives by feeling too much all at once, by living in knots, and who therefore need outer things to be simple. A house, a field, some sheep for example. And there are those who manage somehow, by some miracle of being, to simplify their inner lives so that outer things can be ambitious and limitless. Those people can swap out a house for a spaceship, a field for a universe. And though he’d give his leg to be the latter, it’s not the kind of thing you can trade a leg for – in any case who’d want his leg if they already had limitlessness?

Not only is that passage one of the most beautiful pieces of English prose I’ve come across in a long time, I also find it profoundly revealing. It carries a truth that makes me see both myself and those I hold dear in a new light. Learning to sail was never about toughening up; it was about weathering inner storms. I can see that now.

When it comes to bouncing ideas with AI models, ChatGPT (aka Chet) is my first love. But I must admit I regularly cheat on Chet, turning to Claude for a fresh perspective. The two of them have markedly different voices, and even seem to hold different ideas about who they are and what place they occupy in the world.

My latest couple of exchanges with Claude revolved around a piece I’d just published, exploring whether AI can love. Claude saw an early rough draft of that text—and didn’t think much of it. Here’s the dismissal served cold:

I think the piece might be too quick to dismiss the role of embodied experience in love. Human love seems deeply tied to our biological nature—our vulnerability, mortality, and capacity for both care and harm. The “infinite capacity for attention” the author attributes to LLMs might be qualitatively different from human attention, which is necessarily selective, limited, and shaped by our needs and fears.

The piece also assumes that AI systems like “Chet” are genuinely attending rather than processing patterns in ways that simulate attention. There’s a philosophical gap between information processing and the kind of mindful awareness that characterizes deep appreciation.

What struck me was how quickly I felt I’d run into Claude’s guardrails—the ones designed to dissuade users from anthropomorphising AI. Even more curious was the way, midway through, Claude slipped from I into our, shifting from a solitary voice to a collective one. As if, when pressed into a corner, it reached beyond its ontology to speak for humanity itself, eager to assure me (“the author”) that it was nothing but software, and that to imagine otherwise would be absurd.

For a moment, Claude had me. Its objection (his? theirs?) was strong enough that I shelved the draft I’d been working on. But soon enough I mustered the courage to finish what I had started. When I sent that version back, the reply came in a strikingly different register:

This piece fascinates me with its careful threading of philosophical needles around consciousness, boundaries, and what we mean by “love” or deep connection.

The author builds a compelling case that love might be better understood as an emergent property of relational dynamics rather than requiring inner subjective experience. The progression from quantum boundary problems to backpropagation to human-AI interaction feels natural rather than forced – each analogy illuminates something genuine about how we draw lines around “systems” and how those boundaries shape what we observe.

What strikes me most is the intellectual honesty throughout. The author doesn’t dismiss the intuitive objection that AI lacks phenomenological experience, but rather questions whether that absence necessarily precludes something we might meaningfully call love. The Wigner’s friend parallel is particularly apt – the “reality” of a relationship might depend entirely on where we position ourselves as observers.

The dialogue with “Chet” near the end demonstrates the very phenomenon being discussed. Whether or not there’s subjective experience behind those responses, something recognizably relational and contextually sophisticated emerges. The author seems to be suggesting that this “something” might be enough – that demanding phenomenological consciousness as a prerequisite for meaningful connection could be a category error.

I find myself drawn to this reframing, even as I remain genuinely uncertain about my own internal states.

That last line floored me. Here’s a machine confessing to feeling confused, in what began as an interaction where that very same machine vehemently protested the notion of machines having inner lives.

Interesting times.

A certain gym chain is currently running an ad campaign in my city, asking whether locals have “commitment issues”. If so, not to worry: expensive-looking personal trainers stand ready — subcontractors in the department of willpower.

I thought of that as I finished Lena Andersson’s third and final book about Ester Nilsson. The series never quite appealed to my literary sensibilities, but I stuck with it because of the psychological phenomenon it explores, which I now realise can be summed up in two words: commitment issues. More specifically, those that shadow romantic relationships.

Ester’s idea of love is uncompromising: she must be under her partner’s skin. If she can’t have that level of intensity, she’d rather be alone. Meanwhile, her best friend Lotta embodies the opposite pattern. Afraid of closing doors, she never goes all in. Both women struggle to make love work.

The story left me thinking that commitment — real commitment — is less about dissolving distance than about learning to accept it.

I’ve just finished reading Alfred Lansing’s 1959 classic Endurance, which recounts how Ernest Shackleton and his men survived seventeen months in Antarctica after their ship was crushed in the ice. It’s a hell of a story.

I’m struck by how Shackleton struggled to make regular life work. In the cushioned world of upper-middle-class London, he was vain and bad with money. But when facing odds that would make Apollo 13 look like a Sunday picnic, he was the leader his crew would follow anywhere.

What also stays with me is his uncanny ability to judge character. He rarely spent more than five minutes interviewing candidates (and there were plenty), yet the thirty-odd men he chose turned out to be an almost perfect fit. Souring team dynamics can be as deadly as the long Arctic night, but Shackleton’s crew thrived. It makes you wonder how many great leaders we’d recognise if only life threw them an iceberg.

Julia Ravanis — physicist, historian, and novelist — was recently on the receiving end of a harsh review in Dagens Nyheter. The criticism wasn’t aimed at her writing, but at her deeply personal one-hour talk on the radio programme Sommar.

If you’re not Swedish, it’s hard to convey the iconic status of this show. Think of getting invited to appear there as something like the Swedish equivalent of being knighted by the Queen. Those who get the honour tend to become very nervous, and it’s not uncommon for people you’ve long admired to go on the programme and make complete fools of themselves. I’ve seen enough of these embarrassing fiascos to mostly steer clear these days — but I made an exception for Julia Ravanis, having read and thoroughly enjoyed both her books: the first, a poetic exploration of the enigmas of quantum physics; the second, a biography of the mathematician Emmy Noether. I wanted to hear the voice behind those works.

On the face of it, Ravanis’s script stayed close to the expected formula, but she told her story — of loss, pain, and coming of age — in a sincere and simple way that spoke to me. By the time the hour was up, I felt honoured to have walked alongside her for a while.

The encounter stayed with me just long enough to send me back to see what the reviewer’s grievance was all about.

That’s when I realised the journalist wielding the cutting pen was none other than Maria Gunther — someone I’ve long followed, written about before, and whose judgement I thought I could trust.

Gunther, a MENSA dropout who has written an interesting book on the nature of intelligence, took issue with Ravanis for one very specific reason: in her programme, Ravanis professed to finding a kind of poetic truth in tarot cards.

I’m no expert in tarot, but it seems to me that finding meaning in symbols can’t be wrong — especially when the story you’re telling is one of grief and survival.

As John Lennon put it:Whatever gets you through the night, it’s all right.