This Neural Network Is Generated Some Truly Bizarre Recipes

A neural network was trained to create its own cookbook recipes, resulting in some strange and unappetizing concoctions. 

Do androids dream of cooking? That's the question Tom Brewe asked on GitHub after he successfully trained a neural network to invent its own cookbook recipes. 

Here are the ingredients for a dish called HAWAILIGELED PIE – sounds good, right?

  4 Carrots; finely chopped, drained   -- margarine chopped   1 c Sugar, diced (optional)   1 cn Ice Cream (cleaned)   Spread of fish slices   1 ea Green onions; chopped   2 oz Margarine, coarsely chopped   2 tb Sugar   1/2 c Olive oil   2 sm Eggs

The directions read "Cut into balls. The lettuce in a 10-inch baking sheet and serve with aluminum foil and dice."

The neural network generated other recipes for dishes like BARBARA PULP0ICE, BUFTHA DINGS and CHOCOLATE RANCH BARBECUE

Like the example above, the recipe directions are often just as absurd as the ingredients – one instructs you to beat butter until it's smooth and then drain on both sides of the refrigerator. 

How exactly does the network generate these bizarre ideas? Mostly by trial and error.

Artificial neural networks are computational models that learn by defining associations between things — images, words, letters — similar to how the human brain does. A recurrent neural network, the kind of network that was used to generate the cookbook recipes, analyzes the data it's given (cookbook recipes, in this case) and attempts to mimic it by generating text.

It then looks at what it's already said and uses probability to guess what to say next. The network learns through trial and error, improving with each iteration, like a toddler learning to speak.

"It's constantly making guesses, checking its guesses, refining its own internal neuron connections based on whether it's guessing right or now," said Janelle Shane, a research scientist who runs the blog lewisandquarks.

Shane thought the neural network-generated recipes were hilarious, like the one that called for a half-cup of shredded bourbon. She wanted to see what other kinds of recipes a neural network could conjure.

Someone is teaching a neural network to cook. And now I can't breathe.

— Justin Warren (@jpwarren) March 29, 2017

do androids dream of cooking?

Big Think
Sponsored by Lumina Foundation

Upvote/downvote each of the videos below!

As you vote, keep in mind that we are looking for a winner with the most engaging social venture pitch - an idea you would want to invest in.

Keep reading Show less

Why Lil Dicky made this star-studded Earth Day music video

"Earth" features about 30 of the biggest names in entertainment.

Culture & Religion
  • Lil Dicky is a rapper and comedian who released his debut album in 2015.
  • His new music video, "Earth," features artists such as Justin Bieber, Ariana Grande, Ed Sheehan, Kevin Hart, and Leonardo DiCaprio.
  • All proceeds of the music video will go to environmental causes, Dicky said.
Keep reading Show less

After death, you’re aware that you’ve died, say scientists

Some evidence attributes a certain neurological phenomenon to a near death experience.

Credit: Petr Kratochvil.
Surprising Science

Time of death is considered when a person has gone into cardiac arrest. This is the cessation of the electrical impulse that drive the heartbeat. As a result, the heart locks up. The moment the heart stops is considered time of death. But does death overtake our mind immediately afterward or does it slowly creep in?

Keep reading Show less

Behold, the face of a Neolithic dog

He was a very good boy.

Image source: Historic Environment Scotland
Surprising Science
  • A forensic artist in Scotland has made a hyper realistic model of an ancient dog.
  • It was based on the skull of a dog dug up in Orkney, Scotland, which lived and died 4,000 years ago.
  • The model gives us a glimpse of some of the first dogs humans befriended.
Keep reading Show less