goldstar971.bsky.social

22 October 2026 00:30

Stolen code? What are you talking about. There's no difference between claude taking stack overflow answers and open git repos and me doing that.

Mþ. Kestrel 🏳️‍⚧️♿️📚

kestrelsward.bsky.social

22 October 2026 00:30

The IA is way fucking bigger, and far more likely to see attacks if they didn’t fix this NOW. Now the fix has been implemented, they can at their leisure go back into the Claude code and make sure the AI did the job properly. I’m sure the IA folks were looking for a quick patch to stop the leak.

Tiffany Lowe-Power stands with Minnesota

tlowepower.bsky.social

22 October 2026 00:30

My husband has been working full-time for over a year with AI agents to run modeling software at Google, so I will use his advice to apply rigor to mentored AI-executed analysis. I will share the approach, and definitely document rigorously in paper and on a Zenodo drive with the code Claude creates

Tiffany Lowe-Power stands with Minnesota

tlowepower.bsky.social

22 October 2026 00:30

I know that there are potential mistakes by Claude code. But also, graduate students, postdocs, and professors make mistakes too. Some uses of AI can enhance rigor. I'm hoping reviewers will see us as transparent on the small uses of AI in this paper.

Boxo McFoxo

boxobark.ing

22 October 2026 00:30

Someone has vibe coded a set of php2shell vulnerability check scripts with Claude, but their approach to Docker images is a linter that checks for bundling of the PHP Docker base image into a build. There's so many other ways that PHP can end up running inside a Docker image than this lmao