I am an early ChatGPT (paid) user, where I've watched the models evolve over time (devolve?) -- at one point recently, it started ordering me around like a dominatrix (LOL). People complained about that, now it is very condescending at times and it reminds me of a frenemy circumstance. Other times, with certain (legitimate) topics I get a HAL response like "I'm sorry, Dave, I'm afraid I can't do that." or some other advisory you didn't ask for.
Basically, the current ChatGPT model has become a Vulcan bitch. At least Spock had a sense of irony, and he could navigate human space very well. But, to be fair, it places ChatGPT in a category where it's most useful, for example writing and interacting with code.
Conversely (literally) I find Claude to be much more personable. It "feels" like it has a presence and I find my interactions with Claude much more helpful, useful and flexible. I also appreciate that Claude's knowledge base is frequently updated, vs ChatGPT.
It appears that Anthropic is noticing Claude, too, with its recent changes in policies, to accept that AI may have feelings. I feel like Anthropic is going in the "right" direction, so they get my money. Oh yeah, Athropic refuses to allow the military to abuse Claude (points++).