> whoami
AI is complicated. Explaining it shouldn't be.
50 Shades of AI is a home for long, careful explanations of how AI actually works, plus the projects and resources that go with them. It's for people who want to understand AI, not just use it.
What you'll find here
Long posts that go all the way down, with code you can run and diagrams you can read. Every one is tagged Rookie, Hero or Legend.
ProjectsReal things, built in public, with repos and honest write-ups, including the parts that didn't work.
ResourcesThe papers, courses and tools worth your time, with a note on why each one made the list.
> make issue
How an issue gets made
Every deep dive goes through the same four steps before it's published.
- 1
Research
Read the papers, the code and the arguments on both sides. Primary sources over hot takes.
- 2
Build
Write working code for the idea. If we can't build it, we don't understand it well enough to explain it.
- 3
Draw
Turn the hard parts into diagrams, callouts and examples. Cut anything that doesn't help.
- 4
Review
Check every claim, run every snippet, and fix mistakes publicly when readers find them.
House rules
- No hype. If something is overrated, we say so. If it's genuinely impressive, we show why.
- Show the code. Explanations come with something you can run.
- Cite sources. Claims link to the paper, repo or data they came from.
- Say when we're unsure. "Nobody fully knows yet" is a valid answer in AI.
- Keep posts current. When the field moves, we update the post and note what changed.
