Between me and the robot
Five fragments for Tuesday, May 27, 2025
Hello from Tuesday and the end of May. Next week, a new month—and a new chapter. (Previously.)
Victory: I made it to last week’s Nils Frahm concert after all, albeit delirious from running all day on a mere 90 minutes of airplane sleep. Sleep deprivation somehow made the whole thing even more hypnotic.
Here are five fragments that stuck with me last week…
For more casual, emotional, empathetic, or advice-driven conversations, Claude keeps its tone natural, warm, and empathetic. Claude responds in sentences or paragraphs and should not use lists in chit chat, in casual conversations, or in empathetic or advice-driven conversations.
– Claude Opus 4 system prompt, May 22, 2025. Simon Willison’s gloss is worth a read; as he notes, “That ‘should not use lists in chit chat’ note hints at the fact that LLMs love to answer with lists of things!” Lists: the love language of LLMs.
My first narrative video was with Kopiku Coffee, one of my favorite cafés in the city (shoutout to my friend Helen, who came along as a second camera and moral support). It was definitely a challenge at first — constantly shifting my grip to stabilize the camera (steady videography requires a surprising amount of core strength), viciously churning my conversational wheels during interview lulls, and sidestepping around the café to avoid disrupting the baristas’ flow. Every sudden movement had me scrambling to anticipate the next action, trying my best to re-center the frame to capture every espresso pour and syrup drizzle. But, shot by shot, these awkward movements became a smoother dance, and I found myself settling into a calming rhythm: observe the action, frame the subject, adjust the focus, and repeat.
– Kelsey Wu, “Cringing? Start Anyway,” Working Theories, May 18, 2025. One of the best parts of my past 4.5 years in venture has been the chance to meet people like Kelsey who can do whatever they set their minds to, so imagine my joy at seeing her write in public about one of her side projects: creating mini documentaries for San Francisco businesses. Reading this (and watching her video) inspired me to visit Kopiku Coffee with my daughter the other day…so good.
I want it to start with me and I want it to end with me, and what happens in between is between me and the robot.
– Hilary Gridley on writing with AI, on the How I AI podcast with Claire Vo, May 19, 2025, chock full of tactical tips on how to be better people manager with AI. I listened to this episode while folding laundry and kept getting distracted by how good the ideas were. Not the first time I’ve felt that way about Hilary’s ideas!
In a time when all movies are discussed against a backdrop of the struggling theatrical business, [Tom] Cruise has partially won over film nerds like me because he reliably gets butts in seats. I share his romanticism for the movie house, the communal dream state. When Tom sweats the world sweats with him, and I’m basic enough to cherish that connection.
– David Cole, “I am an audience, first and foremost,” Rats from Rocks, May 24, 2025. I’ve long admired David’s thinking (ever since we were in a book club together circa ’09-11!) and enjoyed reading this essay he deemed “the most ambitious piece of writing I’ve ever done, almost 20k words on Buster Keaton, Tom Cruise, and a hell of a lot of trains.” The structure of the piece delighted me most of all: a 234-word “post” about a tidbit from his hundreds of hours of research, claiming “I’m largely at peace with the ambitious project never seeing light of day,” closing on “Hopefully one day I will write the full post!,” followed by thousands and thousands of words of “footnotes” constituting…the post. Sometimes, it helps to trick ourselves into doing what we’ve always wanted to do.
[While writing code, m]y favorite role is the Detective. You’re in a “Beautiful Mind”-style fugue state and suddenly all the pieces fit together. You experience the elation of successful deduction. You see with perfect clarity why something that should work is broken. You make one careful change in the exact right place and then everything works.
The Genie ends these roles. It does the scientific work of the Doctor, the Mechanic’s work of installing new capabilities, and (less well) the Detective’s work of synthesizing the evidence required to make sense of a mysterious bug.
In their place emerges “the Gambler.” The Gambler does not study the code or make plans. The Gambler keeps asking the Genie for code until something they desire appears in front of them. Then they start asking for something new. And so on until the Gambler gets exhausted, because the Genie never will.
– Drew Harry, “The Gambler and the Genie: How it feels to write code in partnership with a large language model,” March 26, 2025, sent to me by Chris Xu last week. As a history major, I was always fascinated by artifacts capturing people’s “lived experience”—their interior realities, how it felt to be them at that moment in time. This is exactly that sort of document for the present. Drew is balanced about the tradeoffs of this mode; “I don’t feel taxed to my intellectual limits. When I consider a new arc of work, I feel a sense of possibility more than I feel overwhelmed by uncertainty or difficulty.” And yet, taxing ourselves to the limit and emerging triumphant is also one of the best parts of life.
Until next time,
Diana
https://dianaberlin.com