alt Hacker NewsHey, Steve from tldraw here. We use AI tools to develop tldraw. The tools are not the problem, they're just changing the fundamentals (e.g. a well-formed PR is no longer a sign of thoughtful engagement, a large PR shows more effort than a small PR, etc) and accelerating other latent issues in contribution.
About the README etc: we ship an SDK and a lot of people use our source code as docs or a prototyping environment. I think a lot about agents as consumers of the codebase and I want help them navigate the monorepo quickly. That said, I'm not sure if the CONTEXT.md system I made for tldraw is actually that useful... new models are good at finding their way around and I also worry that we don't update them enough. I've found that bad directions are worse than no directions over time.
This is my experience as well. I work with AI agents a lot, they are very useful. What's not useful is some passer-by telling the AI "implement <my favorite feature>" and then sending that as a PR. I could have written a sentence to the LLM too if I wanted to, you aren't really giving me or the project any value by doing that.
Now that writing the code is the easy part, we're just going to transition to having very few contributors, who are needed for their architectural skills, product vision, reasoned thinking, etc, rather than pure code-writing.