I had that before adding the last bit ("bias of your training"), chatgpt is strange, someone else posted a screenahot of a capitalist bias with a similar question just formated differently
I wonder if theres more to it, something about the way it connects make more sense then it being poisioned one way or another. I feel its just something that would be obvious bad to poison.
Its not like adding diversity in images where they van argue their side.
For what its worth, this is the answer/question i get:
I'm curious why this would surprise you. The reasons capitalism, cronyism, and elitist nepotism are dominant is because its the same people who benefit from those systems who would be needed to implement socialism properly. Sure we could and probably will see a populist socialist revolution spread across the world in the coming years. But will the positions of power in such a system be owned by a small group of people or will we distribute decision making across a wide range of people using modern technology? If positions of power are concentrated to an elite group once again (even if that group is initially benevolent and supportive of socialist ideals)... they will eventually take advantage of their power as all prior attempts have seen happen. The system will be tossed off once again in support of a less "late stage" capitalist system controlled once again by moneyed elites.
Openai own and trained chatgpt so poisoning to fit capitalist sentiment, also asking the same question to my self hosted ai has a capitalist bias, I had assumed was due to bias media
It's like crabs (optimized animals) and trains (optimized transportation), optimized economic systems will gravitate towards socialism because it's objectively better in every way that matters, enhancing human flourishing.