Our social feeds are fast becoming awash with posts, podcasts, new headshots, and videos that were generated entirely by AI.
We’re in that first, initial phase of hype and excitement. It’s pretty fun to watch and observe and we are just at the beginning.
But what are the unintended consequences?
Earlier this week I listened to Pack McCormick’s Not Boring podcast, where he makes the argument that the world oscillates between centralization and decentralization. It was great. But, it was not the subject matter that interested me, it was how the podcast was made.
While it sounds like Packy’s voice, the podcast was in fact generated by AI voice, created by the team at play.ht, which used a training model from a bunch of his audio essays.
In Packy’s words, it’s wildly good, and while there are some minor glitches and issues with intonation and inflection, it’s pretty close to being perfect.
I listened to the whole thing, and while it was genuinely mind-blowing, it made me feel uneasy.
Knowing it wasn’t Packy was reassuring, in that he wasn’t trying to pull a fast one, or pass this off as something authentic (I’m not sure it would have passed the sniff test anyway), but it is self-evident that others will not be as honest or transparent in the future.
So what does it mean to use AI to create media and pretend a human generated it?
What are the real-world consequences of this? What should we be talking about to ensure we channel our excitement in the healthiest way possible?
Think about this for a minute.
If you are a copywriter, a journalist, a filmmaker, a grant writer, a ghostwriter, a graduate student, a poet, a novelist, a musician, a playwright, or even a social media manager who has to create ten different versions of the same social post, why are you not using a tool like Jasper.ai to do the hard work for you?
I would. Maybe I am. Maybe this was written by AI. And guess what, I’m not telling. That’s for you to decide in the poll.
Lastly, this moment reminds me of something I was told recently.
Beyond the first movie, did you know that Eddie Murphy didn’t actually do the voice of Donkey because he was so difficult to work with?* Instead, they used an impressionist for Shrek 2-6!
My view of these movies changed in an instant when I learned this, and I would feel the same way if AI was used to recreate the voice of Mike Myers in an animated recreation of Wayne’s World.
It just doesn’t feel right to me.
What do you think? What are the unintended consequences of Generative AI that we need to be talking about here?