Paul then facilitated a modified session of The Thing from the Future where groups created possible future objects linked to the cards they dealt, linked to driverless cars

The key points were:

• Innovation and change can sometimes take a very long time. Artificial Intelligence has been around in one form or another for over 60 years with a couple of hiatuses where it was seen as failed technology and has only got the boost it needed with the advent of using GPUs for machine learning.
• That sometimes innovation does not occur until the technologies to carry out an idea catch up with that idea.
• That practices and business models co-evolve with each other and with technology to form new ways of doing things.
• As ideas, technology, and practice move from genesis to custom built models, and eventually to utility things get easier to use, more certain, and cheaper. Enabling us to build new things on top for less risk and with fewer resources.
• That often the last 5% is harder than the first 95%. Driverless cars have come along way since the first DARPA challenge in 2004 but the “edge cases” are proving just as hard as all the development to date.
• That quite often innovation and new ideas come from left field. Form other industries of other sectors. The Velodyne LiDAR that is used in most driverless vehicles today was invented by a couple of brothers who specialised in high-end audio technology.