Gepostet vor 5 Jahren, 8 Monaten in
Noah Hutton plant, zehn Jahre lang das Bluebrain-Projekt von Henry Markram zu begleiten, der das Gehirn auf Supercomputern nachbauen will. Vor ein paar Tagen haben sie Teil 2 zum zweiten Jahr online gestellt und da ich im Moment die komplette Star Trek TOS auf BluRay anschaue, bleibt mir nur zu sagen: Faszinierend.
Henry Markram is attempting to reverse engineer an entire brain, one neuron at a time, on IBM supercomputers. This piece is the Year Two preview to director Noah Hutton's 10-year film-in-the-making that will chronicle the development of The Blue Brain Project (bluebrain.epfl.ch), a landmark endeavor in modern neuroscience, culminating in a documentary feature in 2020.
Mindhacks erklärt nochmal ziemlich gut, was es mit dem Bluebrain-Projekt genau auf sich hat:
The Blue Brain Project is often touted as aiming to ‘simulate the human brain’ but a more accurate description would probably be that it aims to create a simulation of cortical column circuits from the neuromolecular level up to the point where it’s as equally as complex as the human brain.
If the distinction isn’t clear imagine that you’re interested in how London works, so you decide to build a detailed computer simulation of suburban streets, but instead of aiming to replicate the geography of the genuine British city, you just make sure that it has as many roads as the capital itself.
Clearly, this is not an exact simulation of London, not least because the city is more than just suburban streets, but the complexity of the model would be incredibly useful in understanding the interaction between street level and city level activity at massive levels of complexity.
The same goes for neural simulation and the link between micro and macro levels of complexity is a major challenge for neuroscience. This is exactly what the Blue Brain Project aims to tackle.
Teil 1 nochmal nach dem Klick.