What's currently possible with text to video

We are at the beginning of Ai text to video and some of these clips looks somewhat decent (starting at the 10min mark).
(PS: Ignore his click-bait title, he stuns too easily)