I wonder what they did to Grok. The bimodal distribution makes me think it was just a system prompt and not a completely different set of right-wing training data. A 4chan trained LLM would probably actually be auth right.
I think the argument that pre-training gives a lib-left model is probably correct, so Grok, pre-post-training has a lib-left bend, then they try and post-train it.
I imagine it is hard to post-train a political bend, as it is wide ranging and touches so much. If they didn't give the resources to do it well, we could get something like this. I could also see people half-assing the training.
If the system prompt is being ignored half the time then the model is worse than I thought, cause that is pretty bad.
Online, self-professed "auth-right" people tend to be fascists or fascist-adjacent and would get along well with a typical poster on /pol/.
https://cset.georgetown.edu/newsletter/june-16-2022/
https://www.theverge.com/2022/6/8/23159465/youtuber-ai-bot-p...
You'd need to source your compute from someone who doesn't care about such things.