In case you are wondering, this is actually not me speaking, it’s some AI app pretending to be me while reading my recent review of “Barbie” from Taki’s Magazine. And the AI is doing a pretty good job!
For a few seconds, I confusedly wonder whether I’d recorded this and then had totally forgotten it. It definitely sounds like me, although I’d hopefully read the jokes a little better. It strikes me as slightly more melancholy than I’d sound, but I seldom listen to myself so what do I know about what I sound like?
The world’s leading expert on my speaking style, my wife, is unimpressed, but, then, she’s biased.
And then there’s this, although it gets off to a rocky start, putting an extra S or two at the end of “2020s.”
The app has problems with the term “2020s,” like Senator Liz Warren saying “Latin X.”
The “Oppenheimer” review has fewer jokes, which are hard for the AI to deliver, so it might work better.
The AI is presumably interpreting my prose and reading it with some understanding of what it means. But it doesn’t seem to notice what is supposed to be funny.
I suppose they could build an app where you could annotate the content to explain how to read it, but then I might as well read it myself.
Still, it’s all very amazing vs. a couple of years ago.