As you’ve probably read, this newsletter is getting way more exploratory as I try to understand AI while also drafting a book offline (today is the last day to subscribe for June’s snail mail!)
Here’s everything I explored/read/noticed in May, split by what I’m processing vs. what felt grounding.
still processing:
Started the month processing this video on using AI for structure, production, speed etc., which I wrote about in A Tale of Two Writers. I’ve since read a lot of back and forth on if/when/how writers do/don’t/won’t use AI in writing. The arguments so far feel very immature. I think the people who are being honest about their AI use are brave. But I also see how differently that honesty is received in tech/Substack circles (wow! thanks for the tips!) vs. writing/literary circles (hard no, research is part of writing). I really can’t articulate what I think yet. Ask me again in months.
A dinner by Women+AI—I wanted to go to get a feel for who would be in the room since it is rare I spend much time with people who feel like peer-level folks in the tech world. We chatted about how we are using AI and there was a panel. A few days later, the Girlbossification of AI piece in NY Mag came out and oof do I have thoughts about how flattening this take is—it's a compression I've watched happen in other conversations. It always feels too early to conform, and impossible to have sweeping opinions on something I don't have literacy over yet.
Self gut-check on where I’m feeling comfortable/uncomfortable with AI right now:
Comfortable—how I chat. I’ve ended up developing this weird habit of exhausting a chat until I feel like I’ve processed or built what I need to with it and then deleting it. There’s something nice about the deletion moment that signals to me: this was a tool to help, now I am done, bye. I also keep memory off because I want to be intentional about what I carry forward across conversations. I’ve been surprised to learn a lot of people don’t delete? That said I’m unsurprisingly uncomfortable (read: nervous) about agents (ie handing off full process) but learning how it works because there are processes where this can be really interesting. I just don’t want to give up control yet.
I had a long drive with a Nigerian Uber driver in SF and we chatted about his journey from taxi > uber > doing both. He spends $2700 a month to rent his car from Flywheel, which seems like a lot, but he says he makes $2000 a week driving. He immigrated to SF on December 12, 1990 and studied criminal justice administration here but ended up choosing to drive because it’s the only career that makes enough while giving him 2 months off per year to go back home to Nigeria. I don’t know why he’s still on my mind but he is.
I know a ton of companies are having conversations on architecture right now, and I suspect we need to be having a similar one about life outside of work, but of course, that’s much harder to do. From Databricks: “There is a well-known example from the second industrial revolution. Factories that replaced steam engines with electric ones but kept the same floor layout got almost none of the efficiency gains. The technology changed. The system didn’t. That is exactly where many organizations are right now.”
Very cool to see all this in one place: MIT’s AI Risk Navigator connects MIT's AI Risk Repository datasets through a shared taxonomy.
This quote from Charlie Warzel in The Atlantic pretty much sums up why I felt I needed to get back into reporting: “I’d argue that the most common feeling about AI is somatic: a low-grade hum of difficult-to-place anxiety that’s the result of loud people constantly suggesting that the near future will look very little like the present and that nothing—your job or the social contract—might survive the transition.”
This piece by Elena Verna (growth advisor, now at Lovable) on what she calls the High-Impact Individual Contributor, a senior person who left management to build end-to-end alone, doing in days with AI what used to require a team.
Very good interview from last June with Jasmine Sun and Kevin Roose with interesting points on future of nonfiction books (bleak according to Kevin), why tech hates journalism (poor understanding of conventions, shift in last decade from trade reporting to industry reporting and therefore new interest by mainstream media to prioritize accountability reporting not explainers) and opportunities for form: YouTube + podcasts over writing.
Interesting piece by Lydia Kiesling on “book list slop” from last year. She wrote book previews for The Millions for a number of years and traces how book criticism collapsed into endless lists (The Millions‘ first preview in 2005 had 15 books; LitHub’s 2025 list had 291, written by people who haven’t read what they’re recommending). In this sense, the May 2025 Chicago Sun-Times AI-hallucinated reading list wasn’t a betrayal of the form; it was its logical endpoint.
Molly Kinder, a Brookings senior fellow on AI and the future of work just launched a Substack and I just love it. She argues that more urgent than anything else right now is that we tackle navigating “the messy middle”— the hard period between today’s mostly intact labor market and the post-AGI abundant future Silicon Valley is promising/everyone else is fearing. Here’s her why.
Read Andy Hall's "The Politics of Jobless Prosperity" and this tiny piece stayed with me where he points to an argument from Matthew Yglesias that the fight over data centers is more about NIMBY-ism than AI. There’s something going on here about how maybe we don't understand the thing well enough yet to be mad at the right part of it, so the anger lands on data centers because they're easy and familiar. Not that that’s not angering in itself, but I do think it obfuscates our ability to get curious enough to learn about the technology.
Zilan Qian in Asterisk Magazine on why polling about “optimism” toward AI in China is misread. She argues that the Chinese public has been through 3 cycles now of what anthropologist Xiang Biao calls “last bus” mentality — every shift is the last bus you can catch toward survival: English fluency, mobile internet, AI, and what looks like enthusiasm is fear and experience with mass layoffs plus the absence of a way to refuse the wave.
Newly obsessed with the writing of Audrey Tang (former Digital Minister of Taiwan, free-software developer since age fifteen) who gave an Oxford talk reframing the principles of free software as a care ethics framework for the AI era. Her framework of complementary vs. competitive tools is excellent and also the examples small, local AI helper accountable to a specific community, forkable by it, retirable by it. Example: her family started running one to help her father with medical questions, after ChatGPT kept suggesting fantastical treatments to keep him talking.
Question: Utah just passed the Digital Choice Act (HB 418), the first US law requiring social media platforms to be interoperable — to let users download their data and connections from one platform and transfer them to another, the way email works across providers, which takes effect July 1, 2026. Utah seems unusually active on social media regulation, does anyone know why this is such a thing there? I have done no research on this but curious: is the concentration of Mormon influencers (who, I learned, are sometimes paid by the church) related?
So many takes (too many to link here, most of them polarizing) of people debating AI use in newsrooms and in writing.
This piece (and the comments section) on AI as an accelerant to the already in motion demise of college as it is currently designed was interesting. I’m very interested in conversations about decoupling learning from credentialing.
A new study from the Center for Media Engagement at UT Austin tested whether using AI to rewrite news articles for Gen Z makes them more engaging. Short answer: no but they were surprised by a pattern that many assumed that AI had been used to write the articles even when looking at original, non-AI reporting. I think there is a lot here on how perception and trust works in the new generation and any newsroom dabbling with credibility issues should think about the age of their audience.
The Center for Humane Technology published a piece arguing that we measure AI’s capabilities (bar exam, code, computer use) extensively but barely measure what AI does to us—our minds, thoughts, communities. They’re calling for a new interdisciplinary field of psychosocial AI evaluations and are looking for aligned thinkers.
Anthropic published an updated Economic Index using anonymized Claude conversation data to map which tasks and occupations are showing the heaviest AI usage. None of it was super surprising but it was interesting to see it laid out.
Marija Gavrilov (of Exponential View) shared notes from a recent lecture by Jack Clark, one of Anthropic’s co-founders where he talked about coming back to Anthropic from pat leave and seeing so much change and three rules he called “epistemic hygiene” for the next generation: read the primary material directly, form your own opinion before asking the AI model, maintain independent practices (reading, music, sport, craft) where it is “you versus the world,” minimally mediated by algorithmic systems.
Went down a Theo Baker rabbit hole after watching his Commonwealth Club talk here—he’s a Stanford student journalist (about to graduate) who just published a book about power at Stanford after doing student reporting freshman year that led to the university president’s resignation. In the talk he says he never planned to pursue journalism (he’s the son of two prominent journalists), but I noticed he wrote an op-ed for the Chicago Tribune as a teenager worrying about our lack of controls over AI. Then, this month, as he is about to graduate, another one for NYT about being the first college class to go all four years alongside it. Reading them side by side is weird and interesting: the first feels like teenage nerd noticing something, to college grad reporting on his own generation from the inside.
feeling grounded by:
Friends helped me process why I suddenly feel this immense urgency to do journalism and feel unsure about art, which is exactly the opposite direction in which i’ve been traveling the last few years. where I landed: what if I just do journalism as art?
This project about the bodies buried under the Legion of Honor by artist Quinn Keck, gave me pause. I learned about it through their talk at the San Francisco Center for the Book. I live near and frequent the Legion of Honor, which is a fine art museum and could never quite place why its beautiful campus felt off to me in a strange way. still learning about SF’s complicated histories.
Started a new practice of checking out a stack of books from my own library for each month to nurture the two writers within me
Seeing Hala Alyan’s memoir about fertility shortlisted for a Pulitzer; I think this subject holds so much potential for examining life, migration and systems, as she does masterfully, and yet in most of books that touch this subject, it gets siloed into the world of motherhood.
Listened to Mac Barnett and Jon Klassen speak at City Arts & Lectures about the art of picture books, a genre that isn’t taught the way other literary forms are, but holds so so so much potential for perspective development and joy. also, Mac Barnett just wrote a book on the subject, from which a quote was taken out of context and went viral, but Aroma Umesi does a spectacular job of walking us through the nuances.
Heard Amy Tan speak at Arion Press, the only printer in the US still making books entirely by hand, about 5 books that have shaped her life. It was the most grounding reminder that books don’t just inform us, but form us.
Attended my first cookbook club, a meetup in which participants each cook a recipe from the same cookbook—what a genius, simple, generative way to gather.
This tip from sex educator Kathleen Hema on how to handle big questions (in general) from kids at bedtime: tell them you’ll write it in the question book and discuss it later. Gives you time to think about how to respond and assures them you will make the time to discuss. Made me wonder: what if I did the same for myself, instead of going straight to the AIs?
Came across and loved this summer reading list of translated books!
Happy June!
Jihii