AI and Baking - Is There Potential?

Scrolling through social media lately, I keep stumbling across AI-generated recipes that look incredible at first glance—until you look closer. Some have fundamental structural flaws, like this:

Why is the beef ragu served with both pappardelle and tubular noodles in the same dish, when the recipe calls for only pappardelle?

Others promise impossible results, like this:

How did they pull off this cake?

After going to the actual website (which is here), I’m met with a recipe that is completely different from the picture - zero mention of the crumble topping (instead there is a pourable strawberry glaze), a vanilla cake instead of strawberry, and no attempt to actually layer the cake with the cheesecake.

These AI creations look beautiful, but they simply won't work in a real kitchen. The recipes either violate basic cooking principles or make promises they can't deliver.

But here's the thing: I think there's potential buried in these AI mishaps. What if I took these fantastical recipes and reverse-engineered them into something that actually works? I'm thinking about turning this into a series—taking viral AI baking fails and creating real, tested recipes that deliver on the visual promise.

At the end of the day, you can at least believe that my delicious treats are handmade, no robots involved (except for our cute little Sweet Tooth)! That said, if I were to explore the concept of bringing these AI concepts to life, would you be interested in following along?