I found the conference a fantastic experience with sessions including VR technologies, interactive fabric and gesture-controlled computer interaction.
One the sessions I most enjoyed was 'Machine Learning and Art' from the Google Cultural Institute. Mario Klingemann, who has also previously worked with British Library Labs data, discussed applying machine learning to cultural collections: both for classifying/organising images where metadata is lacking; and computer-generating visual art. Cyril Diagne then talked about visualising/mapping cultural collections using neural networks.
I also attended the annual Women Techmaker's dinner, held the evening before I/O.
The After Hours party was amazing fun with interactive experiments, artworks and games exploring the intersection between code and design (including 3D painting in VR and music visualisation from the Chrome Music Lab).