The 20 best AI sources for 2026, according to my colleague

Over lunch, a colleague and I were talking about how fast AI is moving. Blink and there's another new model, another new tool, another new regulation. I admitted that I sometimes can't see the forest for the trees.

He immediately sent me his own list: the websites and newsletters he uses to catch up every morning. When I clicked through them, I thought it was too good to keep to myself. So here it is, by category, with what each source is good for.

Updated on 25 September 2026. I checked every link and updated where needed. One source no longer exists: Papers with Code (see number 17).

A note for English-speaking readers: four of the twenty sources are Dutch, because the list comes from a Dutch colleague. They're still worth a look if you follow the Dutch tech market.

Tech journalism: for understanding what something means

These are the sources for when you want to look past the hype of the day.

• 1. The Verge (AI): according to my colleague, and now me too, the best place to start. The Verge doesn't just report that something is new, it also tells you what it's like to use.
• 2. MIT Technology Review: sober analysis of what is technically possible. A good antidote when everyone around you is shouting "revolution".
• 3. TechCrunch AI: if you want to know where the money goes, this is where you read about startups, funding rounds and acquisitions.
• 4. Wired: longer pieces, with a lot of attention for ethics and the human side of AI. More of a weekend read than something for in between.
• 5. Ars Technica: for readers who aren't afraid of detail. Technically precise, no clickbait.

Newsletters: the news comes to you

No time to browse? These land in your inbox. The first one especially is worth it.

• 6. The Neuron: daily, light and to the point. It reads like a smart friend catching you up.
• 7. Ben's Bites: strongly focused on practice: freshly launched tools and useful tutorials.
• 8. TLDR AI: free, every weekday, about a five-minute read. A bullet list of links with an emphasis on engineering and research. Scan it and you're done.
• 9. Last Week in AI: if daily emails are too much, this is the weekly alternative, as text and as a podcast. It often goes a bit deeper into the news.
• 10. Platformer: Casey Newton writes about the tech giants, AI and regulation, and often knows what is going on behind the scenes. At least one edition a week is free; the rest is for paying subscribers.

From the Netherlands: regulation and the local market

Because you also want to know what's happening close to home. These four are in Dutch.

• 11. Tweakers: still the place for tech news in the Netherlands. The articles are good, but the real value is often in the discussions underneath.
• 12. AG Connect: aimed at IT professionals. This is where you read how the Dutch government and Dutch companies actually put AI to use.
• 13. Frankwatching: for marketers and content creators. Frankwatching translates the technology into the question: what do you do with it in your content and strategy?
• 14. Emerce: for business decision-makers, with a lot of attention for e-commerce and digital strategy in the Netherlands.

Under the hood: engineering and research

For builders and the curious.

• 15. Hugging Face Blog: Hugging Face is the platform where many open models are shared. The blog covers new open models, datasets and techniques, often with code.
• 16. OpenAI News and the Google DeepMind Blog: sometimes you have to go straight to the source. What can the latest GPT or Gemini model do? Read it from the makers. These are the companies' own announcements, though, so read them as company claims.
• 17. Hugging Face Trending Papers: my colleague's list had Papers with Code here, but Meta shut that site down on 24 July 2025. The old address now redirects to Hugging Face's Trending Papers, which shows which research papers are popular, with their code on GitHub. The detailed per-benchmark leaderboards of Papers with Code did not come back.
• 18. Simon Willison's Weblog: Willison explains very clearly how language models work, what you can build with them and where the security risks are.

Tools and strategy

Finally, two more favourites from the list.

• 19. Futurepedia: a directory of AI tools, organised by category. Useful when you're looking for a tool for a specific task.
• 20. Stratechery: Ben Thompson analyses the strategy of the big tech companies. If you want to understand why a company makes a certain choice, read this. The weekly articles are free; the three Daily Updates a week are for subscribers.

Pick a few, not twenty

Don't try to follow them all at once; you'll never keep it up. Pick one per category that appeals to you. I started with The Neuron and Tweakers myself, and that works fine.

Do you have a source that isn't on this list? Let me know.