21st Century Neural Net Does Astonishing Things To 1950's Cookbooks
The 1950's were a very special culinary time, an unholy blending of convenience foods with "creativity" that yielded all sorts of horrors that continue to fascinate us more than half a century later. This was a period when people made up things like Beef Fudge (contains marshmallow, chocolate chips, and ground beef), Circus Peanut Jello Salad (also contains crushed pineapple and Cool-Whip), and Lemon Lime Salad (also contains cottage cheese, mayonnaise, and horseradish).
A couple of weeks ago, Janelle Shane, a writer and research scientist, wondered what would happen if she took some of these abominations and trained a neural net to imitate them. What amazing things could technology come up with? Shane solicited recipes from her followers on Twitter, who gave her things like the three dishes mentioned above. She assembled them into a dataset, set to work training the neural net, and a week later had a whole slew of technology-enhanced masterpieces!
WHAT HAS OCCURRED CANNOT BE UNDONE
I have trained a neural net on a crowdsourced set of vintage jello-centric recipes
I believe this to possibly be the worst recipe-generating algorithm in existence pic.twitter.com/cwQwOpUNDv
— Janelle Shane (@JanelleCShane) February 7, 2020
There were a few that were virtually indistinguishable from the vintage recipes:
This is one of the AI's more human-seeming recipes (vintage American cooking was EXPERIMENTAL)
Though I have questions about "PG" pic.twitter.com/LRYq46fd7o
— Janelle Shane (@JanelleCShane) February 7, 2020
And then there's this:
What's interesting is how variable the neural net recipes are in their bizarreness.
This recipe is from the same model & generation settings that produced "Jellied Vegetable Salad" and "Potty Training for a Bunny" pic.twitter.com/SChwSEnfu1
— Janelle Shane (@JanelleCShane) February 7, 2020
Technology is fun!
As Shane writes on her website, "Remember that today's AI is much closer in brainpower to an earthworm than to a human. It can pattern-match but doesn't understand what it's doing. Commercial AI is not significantly smarter than this recipe AI. Humans have just hopefully done a better job of preventing it from making oblivious mistakes."