Curation is king — my AI learning resources

Once upon a time I would head to Reddit for answers, either to ask or read solutions others have found, and it was great. Major nerds helping little nerds and most people felt like they wanted to help. Now I find it difficult to understand who is a bot and who is human. In the Attention Economy we exist within, just getting views seems to be what is valued over quality. Thus I have been wondering for myself and others: how can we find humans who value quality and derive more wholesome content for our minds to process and to eventually become part of who we are?

That question of curation has been sitting with me. Then I came across Dylan Castillo’s piece, I hate AI side projects, and this line landed for me:

“I still work on side projects. But now I dread sharing them. The internet is saturated with AI slop. If you don’t have something truly special, it can feel like you’re just adding one more padlock to those sagging, padlock-ridden bridges.” 1

I felt a little dread seeing myself in that. It also gave me another way to think about curation: not only what we choose to let in, but what we choose to let out.

Well, from my trawling of reddit, Bluesky, Youtube and the web in general these are folks who I found to be trustworthy and creating content that is valuable. Some are doing it as an attention funnel, for sure, and others are less on that spectrum. Either way it is the best I’ve found. Please send through your top picks too.


Experience level key

Everyone’s at a different place with tech. I’ve built a little key to help you find content that is more related to your knowledge/entry point.

NumberLevelWhat that looks like
1FollowYou follow along — using everyday tools (Word, email) or a YouTube walkthrough step-by-step. You don’t need to know why it works — just that it does.
2CombineYou combine GUI tools into your own setup — Notion, Gmail, maybe Airtable.
3Vibe CoderYou vibe-code: paste into Claude Code or Codex. You don’t fully understand it, but you can make it work.
4AdaptYou adapt others’ work — clone a GitHub repo, open a terminal, get it running, tweak the code. No GUI — it’s all text.
5CreateYou create from scratch. You write the code, you read the code, you build your own tools.
TitleWhat is it? + LinkExperience LevelHow I rate it out of 7By whom/AboutWhat they offerMy thoughts
One Useful ThingBlog1 → 36Ethan Mollick - professor of Education and Start Ups at University of Pennsylvania - Works with implementing LLMs into education systems.He tests out what is happening in AI and applies it and writes philosophically about its implications whilst staying practical. Also Open Source minded.This blog is the starting point for neutral perspectives on AI and the benefits for implementation in Education. The most balanced human I’ve found on the subject.
Simon Willison’s WeblogBlog2 → 57Simon Willison - Software Engineer - Founder of DjangoHe tests pretty much every big AI release and then shares his learnings as Open Source Projects.Up-to-date application of new tools, with an Open Source bent.
ThursdaiPodcast1 → 45Six co-panelists with different hands-on domain knowledge with AI systemsSix mega nerds with different domain expertise chatting about new releases in AI and use cases — Open Source, AI Labs and etcIn depth weekly discussion on new releases - usually with an interview of devs who made the release.
TetragrammatonPodcast Episode1 → 56Rick Rubin — legendary music producer.
Jack Clark — Co-founder of Anthropic, former journalist.
The podcast episode to start for the creative/philosophically minded individual wanting background on LLMs, that doesn’t need a tech background.Accessible entry point. Non-technical framing from a journalist turned AI co-founder and a producer who thinks in creativity.
AI & I by Every.toPodcast + Blog2 → 56Dan Shipper - Founder of Every.to - serial entrepreneur - Media companyIn depth interviews on one subject matter each time. Usually around using Agents and LLMs to code end to end.Great quality content for specific niches.
Ben BitesNewsletter3 → 55Ben Tossell — exited founder (sold Makerpad to Zapier), investor, can’t code but builds anyway. Started the newsletter six weeks before ChatGPT launched.Talk of what is happening with links to the best new Github repositories and inside word of what is happening.A great quick read in the inbox each week. Like a Thursdai that takes 5 mins to read.
LocalLLAMA subredditSubreddit1 → 54The dark lurkers of RedditGround chatter of what is happening, quick hot takes and increasingly AI crappy click bait.It is a great place to ask questions and get ideas.
WolfBenchBenchmark3 → 56Wolfram Ravenwolf - AI agent evaluator and builderReal-world benchmark for AI agents using 89 diverse tasks. Uses five metrics (Ceiling, Best-of, Average, Worst-of, Solid) to show not just peak performance but consistency and reliability.Essential for understanding which models actually deliver on real agentic work — goes deeper than single-score leaderboards. Helps you choose models for your fallback chain with actual data.
Josh PuckettPersonal Website1 → 57Intentional designer who takes a tailored, elegant approach to designing with AI.Interface.dev < A pay once library to help people design with AI for products they care about.I love his approach and find his emphasis on working for others’ gain helps my system relax and want to be more generous myself.
How I AIYoutube1 → 56By Claire Vo who is essentially running her own SaaS business with the assistance of AI.Chats grounded in ‘show me how you use AI’, so we can watch along and learn what is happening behind the scenes.Great to have a look into how quickly things are moving in AI use, and that no one really actually knows what is happening. It is all an experiment!

Footnotes

  1. Dylan Castillo, “I Hate AI Side Projects,” Dylan Castillo, 20 February 2026, https://dylancastillo.co/posts/ai-side-projects.html.