I can certainly understand the statement. I'm no AI expert, I use the web UI for ChatGPT to have it write little python scripts for me and I couldn't figure out how to use codeium with vs code. I barely know how to use vs code. I'm not old but I work in a pretty traditional industry where we are just beginning to dip our toes into AI but there are still a large amount of reservations into its ability. But I do try to stay current to better understand the tech and see if there are things I could maybe learn to help with my job as a hardware engineer.

When I read about OpenClaw, one of the first things I thought about was having an agent just tear through issue backlogs, translating strings, or all of the TODO lists on open source projects. But then I also thought about how people might get mad at me if I did it under my own name (assuming I could figure out OpenClaw in the first place). While many people are using AI, they want to take credit for the work and at the same time, communities like matplotlib want accountability. An AI agent just tearing through the issue list doesn't add accountability even if it's a real person's account. PRs still need to be reviewed by humans so it's turned a backlog of issues into a backlog of PRs that may or may not even be good. It's like showing up at a community craft fair with a truckload of temu trinkets you bought wholesale. They may be cheap but they probably won't be as good as homemade and it dilutes the hard work that others have put into their product.

It's a very optimistic point of view, I get why the creator thought it would be a good idea, but the soul.md makes it very clear as to why crabby-rathbun acted the way it did. The way I view it, an agent working through issues is going to step on a lot of toes and even if it's nice about it, it's still stepping on toes.

They didn't necessarily say they wanted it to be positive. It reads to me like "chaotic neutral" alignment of the operator. They weren't actively trying to do good or bad, and probably didn't care much either way, it was just for fun.

The experiment would have been ruined by being associated with a human, right up until the human would have been ruined by being associated with the experiment. Makes sense to me.

Anti-AI sentiment is quite extreme. You can easily get death threats if you're associating yourself with AI publicly. I don't use AI at all in open source software, but if I did I'd be really hesitant about it/ in part I don't do it exactly because the reactions are frankly scary.

edit: This is not intended to be AI advocacy, only to point out how extremely polarizing the topic is. I do not find it surprising at all that someone would release a bot like this and not want to be associated. Indeed, that seems to be the case, by all accounts

I think it was a social experiment from the very start, maybe one that is designed to trigger people. Otherwise, I am not sure what's the point of all the profanity and adjustments to make soul.md more offensive and confrontational than the default.