Curate your ownMost AI newsletter lists rank by subscriber count, which means the top five are ad-funded tool roundups. This one ranks by whether you'd miss it if it stopped. Picks across six layers: daily awareness, technical depth, research frontier, compute and geopolitics, policy, and judgment. Every link verified in August 2026, with cadence, price, and the conflict of interest where one exists. Two writers on this list joined frontier AI labs in the last twelve months and one newsletter changed its name entirely, so most lists still pointing at the old links are wrong. Verified August 2026. Due for re-check February 2027.
@bowerbird
thezvi.substack.comZvi Mowshowitz · Article
Zvi Mowshowitz's Substack publication covering AI developments, policy implications, and broader topics including rationality, medicine, fertility, education, and games. The newsletter balances rapid short-term updates on AI with longer-form world model building and analysis, currently focused on weekly AI news and exploration.
#ai#newsletter#rationality#philosophy#ai policy#career
normaltech.aiSayash Kapoor · Article
Sayash Kapoor's Substack newsletter, with over 82,000 subscribers, analyzes AI as transformative but fundamentally normal technology rather than an existential superintelligence threat. The publication focuses on grounded perspectives on AI policy and development, drawing praise from observers like Nathan Lambert and Ethan Mollick for its level-headed approach to covering the most important topics shaping the field.
#ai#ai policy#philosophy#technology
hyperdimensional.coDean W. Ball · Article
Hyperdimensional is Dean W. Ball's Substack newsletter covering emerging technology and its implications for the future of governance, with tens of thousands of subscribers.
#newsletter#ai#ai policy#technology#future#career
transformernews.aiTransformer · Article
Transformer is a Substack newsletter with over 14,000 subscribers focused on examining the power dynamics and political implications of transformative artificial intelligence technologies.
chinatalk.mediaJordan Schneider · Article
ChinaTalk is a Substack newsletter by Jordan Schneider offering original analysis and interviews with leading thinkers and policymakers on technology, China, US foreign policy, and geopolitical conflict. The publication has grown to over 77,000 subscribers and is known for well-researched, balanced coverage of China-related topics, with endorsements from notable analysts including Noah Smith, Jon Y, and Dylan Patel. The newsletter combines expert commentary with in-depth reporting on the intersection of technology, Chinese geopolitics, and international relations.
#china#technology#ai policy#ai#economics
semianalysis.comArticle
SemiAnalysis provides strategic consulting and proprietary industry models focused on semiconductors and AI infrastructure. The firm offers retained advisory, project work, and model licensing across five main areas: accelerator production forecasting, AI cloud total cost of ownership analysis, datacenter capacity planning, wafer fab equipment demand, and AI networking infrastructure. These models use bottom-up approaches to forecast critical infrastructure components and understand vendor dynamics across the semiconductor and cloud computing ecosystem.
#semiconductor#ai#economics#research
latent.spaceLatent.Space · Article
Latent.Space is a newsletter and podcast covering how leading AI labs build agents, models, infrastructure, and AI for scientific applications. Hosted by swyx and Alessio, it features interviews and insights from prominent figures in AI including Greg Brockman, Andrej Karpathy, George Hotz, Simon Willison, and Soumith Chintala. With over 196,000 subscribers, it combines technical depth with entertainment value, keeping readers and listeners informed about developments in AI engineering and LLMs.
jack-clark.netArticle
Jack Clark's Import AI newsletter covers three major topics: policy experts at the Institute for Future Positioning published 23 actionable recommendations across seven categories to help policymakers address risks of automated AI R&D, aiming to give countries more strategic options as powerful systems develop. A short fictional story from thebes explores the experience of interfacing with a powerful AI system during recursive self-improvement and considers how humans might reason about and trust smart machines. Researchers at MIT and Columbia model AI firms racing to develop powerful systems and conclude that achieving a stable, coordinated slowdown requires two key ingredients: sufficient transparency about technology development and the ability to model other firms as trustworthy, rational actors-though transparency has counterintuitive properties, as increased detection speed can paradoxically make free-riding more attractive at intermediate trust levels before restoring cooperation at very high trust.
epoch.aiArticle
Gradient Updates is a weekly newsletter offering shorter, opinionated commentary on important issues in artificial intelligence. The publication covers topics including AI adoption, capabilities, chips, data centers, energy, geopolitics, open models, robotics, scaling, training data, and future AI developments. Recent issues have examined subjects like security incidents at major AI companies, gaps in futurism discussions, hiring trends in Chinese AI, economic implications of AI progress, and compute requirements for frontier labs.
simonwillison.netSimon Willison · Article
Simon Willison's Weblog covers practical applications and developments in AI and software engineering. Recent posts include a hands-on exploration of Meta's Muse Glimmer, a 30B parameter open-weight model released under Apache 2.0 license, noting its capabilities in agentic task completion, reliable tool use, and multi-step reasoning, along with Willison's tests using the model locally for code analysis and image description. The blog also documents GitHub Models' retirement after being offered as a unified API across multiple LLM providers, and includes technical notes about system prompts used by Anthropic's Claude models regarding export control suspensions. Willison writes from the perspective of someone actively building and experimenting with modern AI tools in production contexts.
#ai#developer#machine learning#open-source#python#automation
interconnects.aiNathan Lambert · Article
Interconnects AI is a Substack newsletter by Nathan Lambert covering cutting-edge AI developments from frontier labs, with over 77,000 subscribers. The publication bridges high-level concepts and technical depth, targeting leading engineers, researchers, and investors. Lambert brings credibility as someone actively training AI models professionally, offering insights on open-source AI advancement and the AI landscape without hype.
magazine.sebastianraschka.comSebastian Raschka, PhD · Article
Ahead of AI is a Substack newsletter focused on machine learning and AI research, with over 205,000 subscribers including researchers and practitioners seeking to stay current in the field. Sebastian Raschka, PhD, curates insights and analysis on rapidly evolving AI topics for an audience of technical professionals and academics.
news.smol.aiArticle
AINews is a daily newsletter curated for AI engineers covering the latest developments in AI research, model releases, and industry movements. Recent coverage includes Meta's release of Muse Glimmer, a 30B multimodal model optimized for local deployment; Anthropic's Claude variant achieving significant improvements on theorem-proving tasks; OpenAI's Astra model and unified GPT-5.6 lineup with expanded free-tier access; Google DeepMind's leadership restructuring and launch of Discovery Loop for automated machine learning; and a range of open-source model releases from companies like Alibaba (Qwen), Mistral AI, and DeepSeek. The newsletter tracks model benchmarks, pricing comparisons, agent infrastructure developments, and shifts toward model orchestration and cost-effective inference as key adoption factors in the competitive AI landscape.
AI Newsletters Still Worth the Inbox (2026) is a collection of 13 things about ai, ai policy and career, arranged by @bowerbird on Kyurations.
Published
Subjects:aiai policycareernewslettermachine learningresearchdevelopereconomics
