An update on my favorite newsletters

In Aug 2025 I listed ten good blogs that I enjoy reading. I thought it was a good time for an update. Here is the original list.

10 Good blogs

·

August 25, 2024

For a long time I’ve struggled with news. On the one hand, I like to know what’s going on. I’m interested in politics, policy, global affairs, the economy, etcetera. But as I wrote before here, there is a lot out there that I dislike for lack of clarity and factfulness.

Below is an updated list. Damn I subscribe to a lot of newsletters! These aren’t even all of them. Thanks to writing this post, I was able to unsubscribe from a few. But I have to say, I really love newsletters. When I’m idling, instead of going to Instagram I often first open my email app to see if there are any newsletters to read. When I get interested in a new topic, my first instinct it to try and figure out what the best newsletters are on that topic and subscribe to them. If I am interviewing people for open roles, one of the questions I always ask them is what kind of books or newsletters they read about their field. If a marketing person doesn’t read any marketing newsletters, I’d probably not hire them. Same for data scientists and many other specialists. Reading newsletters is a major hack for learning.

• Experimental History has absolute masterpieces, like the piece about debogging yourself.

• Infinite Scroll is one of the best newsletters there is, full stop. I’ve referenced it countless times in my own posts. It is THE definitive authority on internet culture. Always funny, often infuriating. In the same category, I am also a paid subscriber and happy reader of Max Read.

• How They Make Money produces those great money flow visualizations of listed companies. Just great. Satisfies my financial news desires.

• Benedict Evans Newsletter is snarky, and on point. It’s a bit AI heavy, but then again, that is probably the most important technology change happening right now. The free version is already incredible, but I’m currently a paying member.

• Sustainability by numbers is my go-to for well researched, nuanced discussions on sustainability. Recent interesting learnings for example are on how much electricity and water AI actually uses (not nearly as much as many people, especially Europeans, think).

• MKT1 Newsletter with Emily Kramer is still my favorite marketing newsletter.

• Scott Alexander is just the best way to keep a finger on the pulse of EA / Rationalism (the non-culty side). And of course, his series of San Francisco House Party posts are among the best cultural commentary on SF there is.

• Noahpinion has mostly replaced what I used to get from the Economist. I’m still a fan, but I disagree more than I used to. He is wrong about AI increasing risk of bioterrorism. And his Superintelligence is already here post was… not very good. But the good news is that recently he has posted a string of great, interesting economics posts again.

• Gary Marcus is a staunch LLM skeptic, and he hates everything about the current AI situation. He seems to be quite hurt by the fact that he has been right about a lot of things since 2017, but the whole world mocked him. When GPT-5 came out, he had his moment in the sun with a sort of ‘Marcus was right all along’ moment. The good thing about him is that you usually don’t even have to open the newsletter, as the subject line is enough to understand which event he is going to hate on today. The bad thing is that he is just taking everything too personally. Even Marcus has somewhat quietly acceded that coding agents are good, although he frames that in such a way that he has still been right all along, and that OpenAI, Anthropic et al are finally doing what he has been saying all along. In any case, he is totally right about LLMs. I believe he is right about how AI can progress (symbolic approaches). But he is overall just a bit of a hater.

• Jasmine Sun caught my eye (and everyone else’s, including the NYT where she now contributes regularly) with her posts on Chinese Peptides, and other SF culture stuff. But it was My week with the AI Populists that put me over the hurdle to become a paying subscriber. Almost every post makes me think differently about the world.

• The Pragmatic Engineer is something I’ve been reading and listening to (there is a podcast) more and more. The truth is, I have probably become something of a software engineer myself 😉. Even though I still couldn’t write Javascript manually if my life depended on it, I know enough to understand and take an interest in everything covered. When it comes to agentic engineering1, I actually have even have more experience than some of the guests on the podcasts. Listening to extremely talented old school engineers like Peter Steinberger and Mario Zechner talk about how they work with AI now is both super validating when their methods and conclusions line up with my own, and educational as they invariably have other ideas about some key stuff.

• There are more engineering and deep AI newsletters I subscribe to, which I will summarize in one bullet: Hardcore Software by Steven Sinofsky, Machine Learning Diaries, mandates

• Culpium, by Tim Culpan is the definitive newsletter about the semiconductor industry.

• Owl Posting is an incredibly detailed biotech newsletter. I discovered it when I wrote about the microbiome and mitochondria, and have been following it since.

• Stat Significant is a fun newsletter that does data analysis on movies and music. Why is Nickelback so hated on the internet? Why are almost all new movies sequels nowadays? Always fun casual reading.

• Greg Katz, MD is the best health writer on Substack, full stop. Sometimes I skip a post if its really about ‘old people stuff’, but I read most of them just to learn from the way he thinks. He writes from the frontline, as an active doctor. When I read his newsletters, I also often click into some of the links to other newsletters he mentions and overall learn a lot.

• Numb at the Lodge posts stories that are mostly fictional but rooted in current events. Have to be in the mood for them, but they are very often interesting and thought provoking. The storytelling is always great.

• Ethan Mollick was one of the best AI writers early on. I’ve actually unsubscribed from his newsletter, but I still follow him on LinkedIn. As AI has progressed, it seems he has started to lose touch with the ‘jagged frontier’ of capability, a term that as far as I’m aware, he originally coined. He benchmarks models by letting them create small games or reports. He has started repeatedly making the point that AI capability is increasing faster and faster, which I find hard to agree with at this point.

• Chamath Palihapitiya was great with his collection of ‘what caught my eye’ links. His newsletter was one of the main inspirations for me to start my roundup posts. The others being Noahpinion and Michael Karnjanaprakorn. The simple truth is that I don’t think Chamath is enough of a practitioner anymore, especially in AI. And I also no longer trust him to have as much integrity as I’m looking for.

• Lenny's Newsletter was the reason I originally signed up on Substack. I still subscribe to it, but I rarely read it. My priorities and work have shifted in a different direction. Again, like some of the others here, on AI I find it just a little bit too basic.