This project aims to develop machine learning models and tools that create compelling art and music, and to engender a collaborative community of artists, coders, and machine learning researchers. Its research builds on existing work in image generation using neural networks, and grapples with questions such as how to generate art that is dynamic and surprising, exhibits a long-term narrative arc of sorts, and can be fairly evaluated. In addition, the project more broadly investigates the confluence of art and technology through Artists and Machine Intelligence, a program that brings together artists and engineers to realize projects using machine intelligence and has, since February 23, 2016, blogged about art's symbiotic relationship with technology.
Smart Reply is a deep neural network that, for emails answerable with a quick reply, automatically suggests replies to the email. Someone can edit the chosen reply or send it off as is with a tap. The feature was launched in November 2015 and works by, first, encoding the incoming email into a list of numbers called a 'thought vector'. It homes in on the part of the email most relevant to predicting the reply. Then, based on the 'thought vector', a neural network synthesizes likely short responses to the email, such as 'We'll be there!' and 'I'll see if I can find out.'