As part of London Technology Week, Imperial’s Data Science team invited Dr Sean Hill to discuss his work on the Human Brain Project (HBR). Over the next ten years, the project aims to simulate the human brain using supercomputers and is estimated to cost €1.2 billion.
Dr Hill presented some of the strategies the team intends to use and also highlighted the biggest challenges they anticipate facing. His talk was incredibly thought provoking as it went beyond simply explaining the scientific procedure and included ethical questions around the publication of the material; its utility (or lack of) and also the role of the HBR in wider society.
The project appears to be rooted in biological science at first glance. Yet, on closer observation, it’s mainly about computing. The team aims to collect as much data as possible about the brain, analyse the information and then catalogue it all in one place. Open access simulations will also be built so we can visualize our understanding of the human brain. Whilst this work will enable behavioral theories to be tested, its main strategy is to collect a lot of data and then work towards theory from the collected data.
Gathering the data will raise challenges as the brains used will come from many different animals and much of the research may not be reproducible in humans. As such, there is a lack of trust in the results. Dr. Hill’s suggestion for combating this is a better publication method that is fully transparent on methodology, data and analyses. This would be in contrast to the present system that only presents final papers on each topic. Whilst this is an excellent idea in theory, the reality of actually providing such information could be problematic both in terms of volume of published material as well as the need for additional controls.
Once all of the data has been collected and analysed, actually presenting it will require advances in supercomputing as current hard drives will not be able to operate at functional speeds whilst utilising this level of data. IBM are currently working with the project to develop the computing power but the industry is wary about whether they will be able to profit from this new technology outside of scientific research.
The project itself is very niche and will not even be immediately useful in a clinical situation; however, it will enable pharmaceutical companies to develop medicines for the brain more effectively. Dr Hill explained that the current success rate for brain medication is approximately 50% so drug companies are reluctant to take on such projects. It is hoped that this project will reignite their interest in the field.
The final point raised was the link between this project and wider society: 3% of the budget is for public engagement and this included asking the public early on about their concerns relating to the methodology and direction of the project. Once more research is completed, the results will be published and the public will be asked for their viewpoint again. Alongside this, all of the results will be open access. This begs an important question about the accountability of scientific research to the general public and whether the public has enough expertise to determine the path of such a large technological project.
Overall Dr Hill’s presentation was very effective at explaining key aims of the Human Brain Project. He raised some interesting questions about the ethics of science research and it will be interesting to see how they unfold both within this project as well as within the technology sector as a whole.
The talk ‘The human brain project: Building the brain from big data’ took place on 17 June 2014 at Imperial College London, South Kensington campus.