Not sure about the real programming world, but for sure not in the academic programming one. Admittedly, it's maybe not the same, as the coding is just the tool and not the end goal. The AI is probably a net positive for the physicists who were already good. I'm now using algorithms of CS that i never thought of using, getting to the final objective faster and using a straighter road, but I don't see the AI actually getting there by itself. Let's see what the future brings. I still see it struggling too much with actual reasoning. I don't like to call it reasoning yet as that's not what it is actually doing under the hood. Anyhow. Let's see~~

How long before you think you can tell the AI: go through all the issues on my GitHub project and propose PRs for each one. Make sure to follow the conventions from our codebase and follow the style from @ek and myself. We won't have time to review it so ask the other AI agent to take care of it.