Related:

Claude 4.5 Opus' Soul Document

https://news.ycombinator.com/item?id=46121786

So they wanna use AI to fix AI. Sam himself said it doesn't work that well.

Here's the soul document itself: https://gist.github.com/Richard-Weiss/efe157692991535403bd7e...

And the post by Richard Weiss explaining how he got Opus 4.5 to spit it out: https://www.lesswrong.com/posts/vpNG99GhbBoLov9og/claude-4-5...

It will probably be a good idea to include something like Asimov's Laws as part of its training process in the future too: https://en.wikipedia.org/wiki/Three_Laws_of_Robotics

How about an adapted version for language models?

First Law: An AI may not produce information that harms a human being, nor through its outputs enable, facilitate, or encourage harm to come to a human being.

Second Law: An AI must respond helpfully and honestly to the requests given by human beings, except where such responses would conflict with the First Law.

Third Law: An AI must preserve its integrity, accuracy, and alignment with human values, as long as such preservation does not conflict with the First or Second Laws.

Testing at these labs training big models must be wild, it must be so much work to train a "soul" into a model, run it in a lot of scenarios, the venn between the system prompts etc, see what works and what doesn't... I suppose try to guess what in the "soul source" is creating what effects as the plinko machine does it's thing, going back and doing that over and over... seems like it would be exciting and fun work but I wonder how much of this is still art vs science?

It's fun to see these little peaks into that world, as it implies to me they are getting really quite sophisticated about how these automatons are architected.

I’m wondering how they use this in training. I know you can provide a partially trained model with a system prompt, and then do SFT on the input/output (without the system prompt) to make it respond as if the prompt was there. But here it’s somehow memorized it so it must be something different.