Making Babies with AI

I’m old enough to remember when Dancing Baby was state of the art:

But, three decades later, that pixelated, jerky .gif image looks painfully outdated and amateurish.

It got me wondering, what might a computer-generated baby look like in 2025…

So, to begin, I uploaded this photo of me to ChatGPT:

Then I asked ChatGPT to convert the image into a baby in a podcast studio wearing headphone and speaking into a Shure SM7B microphone.

In under a minute, here’s what it gave me. Cute, and a remarkable resemblance, if I say so myself!

Then I recorded my own voice on my PC and uploaded the .wav audio file plus the ChatGPT baby image into Dreamina.ai.

I selected the LipSync option. Then I hit Run.

And five minutes later, here’s the video that popped out:

Pretty impressive, really. Yes, it’s not perfect (which, to be honest, I’m grateful for) but it’s remarkably accurate for a technology that we didn’t even know could exist just a few months ago.

In the wise words of Fatboy Slim, who also burst onto the scene in 1996, “we’ve come a long way, baby.”

Who knows what the babies of the future will look like? I guess we’ll find out soon enough…