Skyknit: Training neural networks to generate knitting patterns

Knitting is one of my favorite past-times; it is simultaneously meditative and productive. It is also decidedly low-tech and a welcome respite away from the screens that occupy so much of my work day.  The Atlantic is giving The New Yorker a run for its money when it comes to beloved source of long-form journalism and all around excellence (don’t worry, New Yorker, I’ll be ever true)–a friend recently called me attention to a story at the intersection of my professional interests and my hobby.

“Janelle Shane is a humorist who creates and mines her material from neural networks, the form of machine learning that has come to dominate the field of artificial intelligence over the last half-decade…Her latest project, still ongoing, pushes the joke into a new, physical realm. Prodded by a knitter on the knitting forum Ravelry*, Shane trained a type of neural network on a series of over 500 sets of knitting instructions. Then, she generated new instructions, which members of the Ravelry community have actually attempted to knit.” See: Full story from the Atlantic

I fully intend to knit one of Skyknit’s patterns–will post a picture when I do. (However, based on the photos of the Skyknit patterns knit by ravelers, I don’t think we have to worry about needing Sarah Connor quite yet.)

Slightly related–I also came across this site, which collects and transforms traditional knitting patterns through information visualization:

*Not just any knitters on Ravelry, either–once again proving that Lazy, Stupid & Godless is the best forum on Rav.

Leave a Reply

Your email address will not be published. Required fields are marked *