Today (Monday, October 7), the Human Brain Project (HBP) has officially begun. Scientists from 135 institutions met at the Ecole Polytechnique Fédérale de Lausanne in Switzerland to kick off what is described as “the world’s most ambitious neuroscience project”.
The HBP was selected as one of the EU’s flagship research project a full nine months ago. I’m guessing that by Brussels standards, this represents quite rapid progress.
The Human Brain Project’s initial mission is to develop six research platforms: neuro-informatics, brain simulation, high-performance computing, medical informatics, neuro-morphic computing and neuro-robotics.
The neuro-informatics platform will build a map of all the brain’s organizational levels, from the individual cell right up to the entire brain. This will be used to develop the brain simulation platform. The announcement isn’t explicit on this, but HBP leader Henry Markram has publicly stated that what he is trying to do is to build a working model of a human brain.
From 2016, prototypes of the models will be ready for use by scientists around the world. The resources — simulations, databases, etc — will be available on a competitive basis, like large telescopes.
Does this mean that the world’s first artificial general intelligence will arrive sometime around 2016? Probably not. Markram has talked about providing a working model of the brain within ten years. He also adds that “a simulation is not the real thing … it’s a set of mathematical equations that are being executed to re-create a particular phenomenon.”
But if you re-create a particular phenomenon, then surely you re-create it? Of course there are plenty of well-informed people lining up to say that Markram is over-ambitious, and that the money will be wasted. But plenty of other well-informed people think Markram will succeed, although perhaps not exactly within his stated timeframe. What if they are right?