I belive by 2030 or so we will may have models that can generate video, audio, and text that is indistinguishable from a normal human being ("The Zoom Call Turing Test"). But that doesn't mean anything- it's unclear whether such a system is intelligent, or merely a highly generalized mimic.

It will probably require an enormous amount of compute power and data, which means it will only be available to groups with $100+B to spend.