Leading AI models (even Grok) are all a bunch of leftist punks
This title could be clearer and more informative.Try out Clickbait Shieldfor free (5 uses left this month).
A small AI research lab called Unslop.run ran 16 leading LLMs through the Political Compass quiz across thousands of runs and found that virtually all of them consistently land in the libertarian-left quadrant. The only exception is Grok, which splits roughly 50/50 between left and right depending on the run. The researcher, Victor, attributes the liberal lean primarily to training data composition — Reddit and academic writing skew left, and high-quality right-wing equivalents are harder to find. Notably, all models — including both versions of Grok — rejected racial superiority, eugenics, and agreed on several socially progressive positions. The study acknowledges limited scientific rigor but suggests its real value lies in comparing models against each other rather than drawing absolute conclusions.
