One point Joe made in his talk was their use of agile practices not only to reduce cost of changing goals and tooling but also for changing team structures which is key for innovation from my point of view. It is essential to be able to bring new perspectives into a team by changing members since relying on fixed organizational structures may have a direct impact on the outcome which needs to be emergent in an innovation project. He presented an agile on-boarding process where a new team member just get's into the sprint by taking the highest value thing from the backlog which she/he knows how to build or wants to learn about by pairing with someone who has the know how. The team morale induced by agile and a shared purpose are the secret sauce of velocity and bind everything together even when teams change.
Another practice I took home from his talk was how wikispeed deals with the tools needed for pairing and swarming on tasks. As I understood they provide 3 sets of tools two for a pair of developers and one for someone who may help out when the two get stuck, e.g. for demonstration of a technique. The reason behind this is not to interrupt someones work when helping out and frustrate someone by taking over his workplace.
The bottom line of his talk for me is that agile is the way to be for innovative teams since it's all about rapid change in any of the dimensions: team, goal and work. There is a lot to learn from wikispeed how to do this well. Unfortunately the question of purpose will not be as easy to answer for many firms that fail to provide a real value for the society like wikispeed does. Given that agile really yields much free time for by improving efficiency like Joe Justice said in closing his call, an open question is what are the great ideas that bring people together to do something really useful for us and how to turn the ideas into action.
I have stolen the title from a short conversation with Andrea Thomasini on the Thames river boat trip referring back to the HBR article The New New Product Development Game (Takeuchi, Nonaka, 1986) which is describing an iterative, cross-functional team approach for product innovation in the automotive space and is often quoted as the source of scrum.