What to Read
on AI & beyond
A short list of newsletters and audio I recommend when you want to understand AI better, use it well, or make sense of the news. Each entry includes a sample so you can judge the fit.
📰 AI News & Tools
AI News & Tools
Two briefings I recommend for keeping up with models, research, the leading labs and AI policy.
Last Week in AI
by Andrey Kurenkov / Skynet Today
A sourced roundup of AI releases, research, industry and policy, with context that helps distinguish announcements from their implications.
Free newsletter
My take: This is the broad recap I would recommend to a busy reader. It connects model releases with industry and policy, and gives you the sources to inspect the claims yourself.
Sample: Models, research and policy: issue 345 (opens in a new tab)
Reviewed
The Batch
by DeepLearning.AI
A weekly research and industry briefing that explains how new approaches work, what was measured and what remains uncertain.
Free newsletter
My take: I recommend The Batch for research explanations that make the details understandable. Andrew Ng's opening letter has a strong point of view, and DeepLearning.AI also sells courses.
Sample: Research and industry briefing: issue 373 (opens in a new tab)
Reviewed
🔍 AI Deep Dives & Workflows
AI Deep Dives & Workflows
For understanding how models and agents work, with technical explanations, experiments and practical implementations.
Simon Willison’s Newsletter
by Simon Willison
Practical investigations of models, agents and tools, with working examples, reported costs and attention to security and failure cases.
Free newsletter; separate paid monthly digest
My take: Simon shows his work: what he tried, what it cost, and where it failed. I recommend him to anyone who wants to understand the systems beneath the interface.
Sample: Testing Claude Sonnet 5.5 (opens in a new tab)
Reviewed
Ahead of AI
by Sebastian Raschka
Technical explanations of models and agents, combining research comparisons with code and implementations you can inspect.
Free and paid articles
My take: If you want to understand how models and agents work, Raschka explains the components and walks through real implementations. Expect technical detail and examples you can inspect.
Sample: Components of a coding agent (opens in a new tab)
Reviewed
Interconnects
by Nathan Lambert
Technical analysis of model training, post-training and competition between AI labs, from a researcher who has worked on open models.
Free and paid articles
My take: I recommend Lambert for the technical reasoning behind model progress and lab competition. His experience training models gives the analysis substance; his advocacy for open models is part of the perspective.
Sample: The balance of power in open models (opens in a new tab)
Reviewed
💡 AI, Society & Big Ideas
AI, Society & Big Ideas
Research, experiments and audio essays I recommend for thinking through AI capabilities and their consequences for work and society.
One Useful Thing
by Ethan Mollick
Experiments and research on how AI changes work, with concrete examples of what current systems can do and where human judgment matters.
Free newsletter
My take: I recommend Mollick for ambitious experiments and questions about how work changes. His optimism is part of the perspective; I would treat the broader predictions as arguments to test.
Sample: The dot and the swarm (opens in a new tab)
Reviewed
AI as Normal Technology
by Arvind Narayanan & Sayash Kapoor
Research and arguments about AI capabilities, adoption and policy, with methods and limitations explained so you can judge the evidence.
Free newsletter
My take: I recommend this for testing AI claims against evidence. The authors explain their methods and limits, including where their own scepticism could affect the interpretation.
Sample: Testing agents on open-ended tasks (opens in a new tab)
Reviewed
The Framework
by Marcel Salathé
Audio lessons and ongoing explorations on how AI works and what it changes for work, education and science. English audio for iOS and Android.
Five free listens, then a paid subscription
My take: Marcel is someone I trust on AI. The ongoing explorations are the part I recommend for readers who already know the basics and want to think through the implications.
Sample: Marcel on five myths about AI and the app (opens in a new tab)
Reviewed
🎨 Product & Design
Product & Design
For the judgment, incentives and coordination that matter when AI changes how teams build products.
The Beautiful Mess
by John Cutler
Product and organisation thinking about judgment, incentives and coordination, including what changes when AI makes production faster.
Free newsletter; optional paid support
My take: I value Cutler for the organisational consequences of AI: what speed does to judgment, feedback and coordination. The practical question is how to improve the work as well as automate it.
Sample: AI and the loss of positive friction (opens in a new tab)
Reviewed
🚀 Entrepreneurship & Business
Entrepreneurship & Business
For understanding AI business models, distribution and who captures value.
Stratechery
by Ben Thompson
Analysis of AI business models, distribution and competitive power, grounded in how technology markets work.
Free essays; paid updates
My take: For AI strategy, I recommend Thompson's analysis of distribution, incentives and who captures value. I would treat his predictions as a coherent thesis to challenge.
Sample: Apps, agents and aggregation (opens in a new tab)
Reviewed
Stay in the loop
Don't miss what's next
One email when I've learned something worth sharing, roughly once a month.