I have been trying out Claude Code a bit since reading all the hype about how good it is. It does work pretty well, but one thing that was a bit worrying was that every time I started a new session, it asked me to allow Anthropic to train on my code. I had already opted-out when that was first introduced, so it should have been redundant at that point.

I don't want them training on my code, and it makes me not want to use it at all seeing how keen they are to do it. From opinions I read online, people seem more than happy to allow AI to train on their code, and I really don't understand why. Do you not value your work? I highly doubt that companies like Anthropic, Google, or OpenAI are freely adding all their IP to the training data. Is it because you think these big companies are going to use it for good?

People like to bring up open source and how it is done for the good of the group, but I don't believe open source code is created out of pure benevolence. There is some value derived from having your name attached as creator of whatever open source code. The only person I would put an exception for is Satoshi Nakamoto, but that guy is either dead or a saint.