Artificial intelligence is fueling artistic experimentation from painting to music. Now creatives can spend less time keeping up with the expanding AI palette and more time making digital masterpieces.
What’s new: Runway ML is a desktop clearinghouse that aims to connect visual artists with machine learning models that ignite their imagination and sharpen their productions. Browse the Image Generation category, for instance, and you’ll find a GAN that can make a selfie look like a Monet, or a semantic image synthesizer that transforms Microsoft Paint-quality doodles into photorealistic pictures.
How it works: Creators of AI models can upload their projects via Github.
- Developers retain rights to their software, but don’t expect compensation.
- A Netflix-like user interface organizes models according to their output.
- Users can run models in an easy-to-use workspace.
- The app, which is in beta, charges $0.05 per minute of compute and comes with $10.00 in free credit.
Behind the news: RunwayML began as a thesis project when artist and coder Cristóbal Valenzuela was a graduate student at NYU’s Tisch School of the Arts, according to The Verge. Now he works on it full-time.
Why it matters: AI opens a universe of creative possibilities: processing existing works, generating new ones, mashing up styles, creating new forms of order out of chaos and new forms of chaos out of order. Runway gives artists a playground for experimenting with these emerging resources.
We’re thinking: Making models accessible to artists doesn’t just broaden the market for AI. Engineers who listen to artists’ creative needs may find themselves taking on challenges they wouldn’t imagine on their own. Like, say, deciding what is and isn't art.