Formal setting
A lobby in a favourite hotel.
Quote of the Day
”Once, when a British Prime Minister sneezed, men half a world away would blow their noses. Now when a British Prime Minister sneezes nobody else will even say ‘Bless You’.”
- Bernard Levin
Came to mind while watching the Trump-in-London circus last week.
Musical alternative to the morning’s radio news
Jerry Garcia | Deal
Long Read of the Day
are you high-agency or an NPC?
Fabulous piece of onsite anthropology by Jasmine Sun in San Francisco on how Silicon Valley’s AI boom has created a new social hierarchy based on perceived human agency and adaptability, sparking both swagger and anxiety among tech workers.
As Meghan O’Gieblyn writes in God, Human, Animal, Machine, human exceptionalism is a stubborn beast. We prize ourselves not on a fixed set of traits but on having whatever other beings don’t. For ages, smarts were what separated man from his fellow mammal. Cheetahs may have speed and chimpanzees strength, but inventing fire and writing was what put humans on top.
Now, LLMs are toppling traditional intelligence benchmarks one by one: the Turing Test, then the LSAT, then the IMO Gold. They can answer PhD-level economics questions and creative writing prompts. But today’s computer-use agents can barely share a Google Doc without human intervention. LLMs can draft an essay pitch but not come up with the concept, give you a recipe for a bioweapon but not the savvy to acquire the ingredients. If agency combines autonomy (“the capacity to formulate goals in life”) plus efficacy (“the ability and willingness to pursue those goals”), AI in 2025 is sorely lacking in both.2
It turns out the secret of human civilization was not any particular cognitive creation but our unending flexibility. To hit a wall and build a ladder to climb it, to design cars instead of faster horses, to come up with new levels of Maslow’s hierarchy to summit once we’ve satisfied the first.
For now, agency is still a human moat.
Do read it for a riveting picture of a segment of society that has become untethered from reality.
The ‘three-sickbag spectacle’: tech bros at the court of Trump 2.0
My latest Observer column…
If you’re as puzzled as I am by what’s going on at the moment, then Antonio Gramsci, the great Italian Marxist philosopher and political theorist, is your man. He spent 11 years in jail at Mussolini’s pleasure, during which time he wrote a remarkable series of “prison notebooks” containing his reflections on Italian politics, culture and history, as well as broader considerations of ideology, hegemony and revolutionary strategy.
There’s a passage in Notebook 3 that seems particularly relevant now. It’s about what Gramsci called an “interregnum” – a period of crisis “which consists precisely in the fact that the old is dying and the new cannot be born; in this interregnum a great variety of morbid symptoms appear”.
Eight months into Trump 2.0, it’s pretty clear that we in the west are deep into an interregnum of our own, and morbid symptoms are everywhere on view. Here are a few, in no particular order, for your leisurely contemplation.
Exhibit A is the dinner Donald Trump hosted at the White House for the tech titans of Silicon Valley. The assembled masters of the universe (Mark Zuckerberg, Satya Nadella, Sam Altman, Sundar Pichai, Bill Gates, Greg Brockman, Sergey Brin, Safra Catz, Tim Cook and sundry others) were there for obvious purposes: to pay homage to the supreme leader, to laud his wisdom and to boast about how much money they were preparing to invest to Make (His) America Great Again. The deeper message was that the most powerful corporations in the world are signed up to Trump 2.0. It made for a three-sickbag spectacle of which the only redeeming defect was that Elon Musk was nowhere to be seen…
Linkblog
How an old newspaper in Hiroshima is keeping the memory of survivors alive Link
Hiroshima 11:00 on 6 August 1945. Photographer: Yoshito Shigematsu
A reminder that journalism is sometimes the first draft of history. The full archive is here.
Feedback
My recent Observer column about AI and the em-dash controversy, prompted Steven Leighton to point me to an intriguing essay on the subject by a former proofreader, Nitsuh Abebe, in the New York Times Magazine.
Large language models are trained on whole mountains of human-generated prose, including far more old printed matter than you or I will ever absorb. We humans ask them to mimic our writing, but we do not always specify — may not even realize — that what we mean by “writing” now includes the practically oral communication we lob through our screens all day. Then we scan the results, find telltale traces of books and magazines, and begin to fixate on those artifacts as faintly robotic. The machines are vacuously reflecting our own traditions back at us. What we may not realize yet is that we are sliding toward new ones.
This Blog is also available as an email three days a week. If you think that might suit you better, why not subscribe? One email on Mondays, Wednesdays and Fridays delivered to your inbox at 5am UK time. It’s free, and you can always unsubscribe if you conclude your inbox is full enough already!