It's interesting how in Star Trek they had a holodeck full of completely believable visuals, amazing generated sound, "hard light" making believable surfaces, characters who were extremely realistic, etc. But, the only holodeck character to ever be considered possibly sentient was Moriarty, and it required that they ask for a character capable of outwitting Data. Meanwhile, the sentient "computer" in the series was Data, who had a lot of quirky mannerisms that made it obvious he wasn't at all human, despite really wanting to be.
In the real world, Silicon Valley is convinced that because GenAI can make something that can mimic a human they think they're weeks or days away from AGI which will lead immediately to ASI. Meanwhile, nobody in Star Trek thought that holodeck tech was a profitable area of research for true AI. I think I trust the Star Trek writers more than Silicon Valley on this one.