Loading…
This event has ended. View the official site or create your own event → Check it out
This event has ended. Create your own
View analytic
Friday, April 29 • 5:30pm - 6:00pm
Making Weird Things With Deep Machine Learning

Sign up or log in to save this to your schedule and see who's attending!

Over the last few years, the world of machine learning has been turned upside down by recurrent neural networks and so-called "deep learning".

Research applications aside, it turns out that these sorts of neural networks are incredibly good at creating generative art. From Google's "Deep Dream" to Twitter bots like @RoboRosewater and @DeepForger, people are starting to use deep learning to create procedurally-generated artworks with far more complex and beautiful output than traditional procedural content generation techniques can produce. Machine learning seems scary and inaccessible, but in actuality there are a lot of high-level tools that make it easy to train your own neural networks. You too can make cool and weird things with machine learning, even with zero background knowledge!

This talk will start with a (very high-level, aggressively approachable) conceptual overview of what machine learning is, and quickly move into a more practical discussion of how game designers and creative coders can use existing deep learning libraries/tools to do all sorts of different kinds of procedural generation, as well as the relative strengths and weaknesses of neural networks as opposed to other forms of procedural generation.

Speakers
avatar for Mike Lazer-Walker

Mike Lazer-Walker

Playful Systems, MIT Media Lab
Let's talk interactive fiction! Or interaction design! Or functional programming! Or hardware!


Friday April 29, 2016 5:30pm - 6:00pm
Redstone Theater

Attendees (23)