Since Musk bought it? Yes-ish. Per Business of Apps, Twitter has maintained around the same user counts, but that is through gains and losses across demographics. Women have fled the platform; what used to be fairly equal is now 2:1 men to women. US users are down 15%-20%. What has grown commensurately is right wing international users, further skewing the average political slant on the platform.

Honestly I wonder if this skews the data at all. 10 people use Twitter, 1 of which is fascist; 5 decent people leave, and now the site is 10% more fascist. Is that driving users more rightward, or driving off more left audiences?

Mmm, I mean, sentience is a gradient, right? The mirror test is where we decided to draw the line, but there are more places to do so. My toddler thinks his favorite toy has some level of agency, just as by all accounts his older sister thinks Bluey has an identity. Depending on the test, there are developmental markers where we statistically transition from failing to succeeding. Another way to look at it is that for each developmental range, we can develop tests that challenge how we perceive autonomy, which some people succeed at and others fail. We may have just inadvertently developed a test that a significant amount of adults are just going to fail as human beings.