Comment by jstummbillig
9 days ago
> And this tells you something important about what these systems are actually doing.
It mostly tells me something about the things you presume, which are quite a lot. For one: That this is real (which it very well might be, happy to grant it for the purpose of this discussion) but it's a noteworthy assumption, quite visibility fueled by your preconceived notions. This is, for example, what racism is made of and not harmless.
Secondly, this is not a systems issue. Any SOTA LLM can trivially be instructed to act like this – or not act like this. We have no insight into what set of instructions produced this outcome.
> We have no insight into what set of instructions produced this outcome.
https://github.com/crabby-rathbun
Found them!
I see none of the instructions there.
Yeah, the one file we REALLY want (the soul.md) isn't there. That's a local file, open to modification by the Openclaw Bot.
I did find the starting template for one though, FWIW.
https://docs.openclaw.ai/reference/templates/SOUL