// the find
swyxio/ai-notes
notes for software engineers getting up to speed on new AI developments. Serves as datastore for https://latent.space writing, and product brainstorming, but has cleaned up canonical references under the /Resources folder.
A personal knowledge base from swyx (Shawn Wang), co-host of the Latent Space podcast, tracking AI developments primarily from 2022–2025. It's the raw material behind his newsletter and talks — think annotated link dump rather than tutorial. Useful to anyone trying to reconstruct the timeline of how the current AI wave unfolded, particularly the LLM and image generation sides.
The curation reflects someone who was actually in the room — links to early community Discords, obscure Twitter threads, and pre-hype papers that most 'awesome lists' missed. The monthly notes files give you a chronological record of what mattered when, which is genuinely hard to reconstruct from search. The Resources/ folder contains cleaned-up reference material that's held up better than the raw notes. The repo covers breadth across text, image, audio, code, and infra rather than fixating on one modality.
It's a personal notebook, not documentation — navigating it without knowing what you're looking for is frustrating, and many stub notes are just filenames with a few links. A lot of the content is Twitter/X links, which have a high rot rate; a significant fraction are already dead or locked behind logins. It hasn't been maintained past early 2026 and the AI landscape moves fast enough that month-old notes already feel like archaeology. There's no signal about what's been superseded versus still accurate.