In 2010, world-renowned Israeli neuroscientist Henry Markram produced a dazzling video promoting his Blue Brain project — a plan to create a computer simulation of a mouse brain which would be the first step in simulating a human brain within 10 years. It opened with Johann Strauss’s “Blue Danube” waltz.
Catchy perhaps, but maybe such an allusion to Stanley Kubrick’s “2001: A Space Odyssey” (1968), with its ill-fated, superhuman computer HAL 9000, was not the best way to introduce a potentially paradigm-shifting development in AI research.
But if anyone were troubled by this ominous reference, Markram’s confidence and charisma, his reputation as a visionary and a genius, would most likely have swept away doubts. Certainly, 22-year-old filmmaker Noah Hutton was convinced enough to commit 10 years of his life to filming the pioneering researcher’s ambitious, if not hubristic, quest.
The result, “In Silico” (the title refers to the potential of studying the brain in a computer simulation, as opposed to “in vivo” with a living organism or “in vitro” with a specimen in a test tube or petri dish), offers a stimulating glimpse into the mechanics and mysteries of the human brain. It also shows the scientific process in all its messy glory, with its revelations, missteps, labor, frustrations, self-promotion, fights for funding, and fractious egos. And it shares a filmmaker’s progress from an awestruck acolyte to a wiser and more objective observer.
The Blue Brain project takes its name from the IBM Blue Gene computer it uses, a descendant of the Deep Blue model which had defeated chess grandmaster Garry Kasparov in 1997. At first, it makes impressive strides. In 2011, year two of the project, the mouse-brain simulation shows evidence of neurological activity independent of what had been programmed into it.
Time to shout, “It’s alive!”? To ascertain the magnitude of this first big step Hutton speaks to experts who are not convinced by Markram’s claims. They insist that he has far too little available data to construct a viable, let alone autonomous and functioning computer model of a living brain. But Markram ascribes such nay-saying to outmoded thinking and staid scientists fearful of losing their relevance.
More important, those with deep pockets are impressed. In 2013, in a competition with several other proposals, Blue Brain is one of two winners of a European Union grant of one billion euros.
But the money, the subsequent vast expansion in resources, and the participation of scores of new scientists bring unexpected problems. Chief among these is Markram himself.
At first it was personal circumstances as well as scientific idealism that drove Markram in his quest. His son suffers from severe autism, a condition which struck Markram as a reproach as well as a tragedy. “I felt powerless,” he says. “As a neuroscientist I could not even explain what was happening to my son.” A brain computer simulation, he was sure, would disclose the cause of not only autism but also Alzheimer’s, schizophrenia, and other insoluble disorders.
But the new funding and the subsequent increase in his power and in the scope of his enterprise exacerbate some of Markram’s less appealing qualities. Colleagues bristle at his authoritarian approach and his refusal to countenance differing opinions. Some wonder if his single-mindedness verges on megalomania, a kind of “Messiah complex” as one critic puts it. Also troubling is the public sourcing of the funding which, as was the case with a similar brain project in the United States, might be drawn from military or other dubious programs. Then there are basic ethical questions: How would this project affect human nature in the long run? Who would control its potential world-changing applications? Markram does not seem to take these doubts seriously.
At this point Hutton begins to look at Markram and his Blue Brain project with a more critical eye. The visual models still dazzle — they are like illuminated, ceaselessly metamorphosing, 3-D virtual Jackson Pollock canvases. But do they have any meaning beyond the gleaming laboratories in which they have been created? When “The Blue Danube” waltz plays on the soundtrack again, near the end of the film, it does not suggest a pending, catastrophic outcome of overreaching science, but rather an intricate, elegant, and beguiling artifice.
Be that as it may, the project resolutely continues. Its new date for simulating a human brain is 2050.
“In Silico” can be streamed via the Coolidge Corner Theatre’s Virtual Screening Room, beginning April 29. A Science on Screen virtual panel discussion can be streamed on April 29 at 8 p.m., with director Noah Hutton in conversation with director Werner Herzog, moderated by Molly Webster of Radiolab. Go to coolidge.org/films/silico.
Peter Keough can be reached at firstname.lastname@example.org.