I’d argue that learning about the effectiveness of AI models and techniques, benefits from voices with different contexts, as well as voices that may be differently aligned with bull vs. bear cases of macro AI strategy.

And those voices may be unequally represented in X for similar reasons - perhaps (somewhat) uncorrelated from politics, but simply due to the UX consequences of prioritizing commenters willing to pay the platform.