John Wood of UCL and former chair of ESFRI, compared his experiences of Cambridge – data gathering at the lab all day, write up the paper with a colleague over wine in the evening – with those of his son, who can access facilities around the world without leaving his room. Professor Wood predicts the emergence of the data scientist – someone who can cope with hexabytes of data, counter data terrorism and tackle data provenance issues. How do you know whether to trust the data you’re getting from a dozen different sources?
Christian Joachim of CNRS in France talked about reversing what he called the miniaturisation process – instead of breaking down a complex system (like the brain) to try to understand its parts, what happens if you build one up, atom by atom? Does it look like the brain, or do you get something completely different?
Dieter Fellner, of the Fraunhofer Institute in Germany introduced the challenge of data handling in a knowledge society. Object oriented computing will lead to applications as we understand them disappearing. Since the 90s, digital libraries have stored text in a way that is searchable and retrievable. We still do not yet handle other media types and 3D objects as well as we handle text – at the moment we still store this computer-born data as pixels, then re-render it. The semantics should be stored with the data, currently we now leave it in the user interface. We should work with these types of data more like we handle music – you search for what you want, listen to it, and download it.
Roberto Saracco of Telecom Italia reminded us that moving from ideas to innovation takes time, money and regulation. You couldn’t justify developing the brain just because it would move your little finger, you need to get the big picture. Businesses will start to think less about selling stuff, and more about the services that go with the stuff – in the future you won’t be able to buy a washing machine without getting a service along with it.
For Martin Curley of Intel Labs in Ireland, the key themes for the future are digital transformations, sustainability and mass collaboration. You can now get a cloud computer on a single chip. For exascale computing you need billion way parallelism and this is 10,000 times away from where we are now. There is also a move to try to make innovation in technology more predictable and hence more profitable – for example, film photography, one of the industry’s biggest money makers is now officially dead with the growth in digital photography. But this year, more photos are taken by mobiles than digital cameras, so is the digital camera itself already on the way out?
Zoran Stancic, European Commission, DG-INFSO had the final word. The EC is interested in using pervasive ICT innovation as an enabler of science and technology. E-Infrastructures, for example, go way beyond normal research infrastructure- they can support the flow of ideas and generate new ones. Establishing European innovation partnerships is also important. In the context of the economic crisis, when we think about boosting support for innovation, we must also support science-driven research to meet long term as well as short term aims.
Asked to think about e-Infrastructures in particular, the panel encouraged us to think about the democratisation of resources – for example what is the impact of the Square Kilometre Array on the local population in Australia or South Africa which hasn’t yet seen a broadband connection? Investment by countries should keep pace with their ability to take part in the infrastructure.