AI Is Being Used To Create Terrifying Masks And Costumes For Halloween

Happy Halloween! For today’s discovery, I found some interesting ways in which artificial intelligence technology and machine learning were used to make Halloween an even scarier time of the year.


Matt Reed, a creative technologist at a Nashville-based advertising agency, trained a neural network by supplying it with 5,000 images of popular Halloween masks. The AI used a general adversarial network, which is a type of algorithm in automated machine learning software.

According to Reed, “it works by pitting two networks against each other allowing them to continuously improve over time. It basically teaches itself how to make itself better. The longer it runs, the ‘smarter’ it gets.”

It only took 24 hours before the algorithm created terrifying monster masks. Reed is currently running the algorithm until the end of October and features the latest creations in a Dropbox folder. 


Additionally, Janelle Shane, a research scientist who blogs about neural networks, has experimented in the past with neural nets to generate recipes and headlines. Now, Shane is using her AI to generate Halloween costume ideas. You can snag a full list of them here.

Uncanny Music Box

Furthermore, Pinar Yanardag and her team worked on AI while she was a postdoctoral associate at MIT Media Lab. This particular AI can generate scary songs for Halloween. Yanardag primed the AI with 5 to 10 seconds of horror movie soundtracks and had it generate new melodies in response. You can check out their uncanny music box application here.

Botnik Give Yourself Goofbumps

Finally, Botnik is a community of writers, artists, and developers using machines to create things on and off the internet. Botnik has a story called ‘Welcome To Sand Hands’ which uses AI-written text to tell a creepy story that allows you to choose from 20 different scary endings. You can check out the story here.

For more detailed articles on these topics, check out TechXplore, Popular Mechanics, and Daily Mail.

«Back to Blogs & Articles

Like the post? Share it:
Andre Moncayo
Andre Moncayo
Marketing Associate