I recently watched the video the New York Times produced demonstrating a real voice actor (the great Hank Azaria) against AI versions of him.
It was fascinating.
But also...kind of offensive.
Lemme explain.
The jist of this video is that AI is pretty good but it can't nail a perfect human voice performance.
This is evident in the part of the video where Hank feeds his 'Mo' voice into an AI audio software. Though the 'machine' can get the tone of the 'Mo' voice it doesn't pick up the nuances of the voice. Plus, it doesn't say words like Mo does like 'dere' instead of the written 'there.'
So, the point of the video is that AI isn't that good. That human voices are better.
And here's where it kind of pisses me off.
The New York Times, who produced the video, has audio versions of it's articles narraed by, wait for it, ARTIFICIAL INTELLIGENCE!
Now, I'll give you that an audio reading of a news article doesn't require the tremendous nuance of a Hank Azaria character from The Simpsons, but it still stinks of hypocrisy.
If you're producing videos about how human voices are better, why are you still using AI to narrate your news articles?
These are articles that SHOULD be narrated by actual humans like me who do this for a living. Why?
- There's more depth to the narration.
- It feels more 'lived in.'
- The power of a passage lands with more impact than a machine reading.
- And, there's a bunch of actors who do this for their livelihood!
Would they do a video showing an AI writing an article for the Times?
DOUBT IT!
Either don't do these videos that make you look like hypocrites or pay actors to narrate your articles.
You can't do both!