Building a Human Brain
Could supercomputers create an artificial brain that can learn new behavior and develop cognitive skills? Some scientists say not if, but when
Last week I wrote about scientists thinking big. And they are thinking big. But compared to Henry Markram, they’re idea lilliputians.
His dream is to build a human brain. Not a real brain of tissue and blood vessels and neurons–but the ultimate super computer, an enormously sophisticated model that would function like a brain, able to learn new behavior and develop cognitive skills. It would be, he says, ”the Hubble telescope for the brain.”
Markram, a neuroscientist at the Swiss Federal Institute of Technology, has been on this track for a while, at least back to mid-1990s. But his quest picked up steam in 2005 when he convinced the institute to invest in a “Blue Gene” IBM supercomputer, and then launched what he called the Blue Brain Project, his first big step in having a machine that could simulate brain functions.
Now he wants the European Union to go all in with him. His project is one of six finalists for its two “Flagship” initiatives. The prize? An investment of 1 billion euros or roughly $1.3 billion.
So what would the EU get for this tidy chunk of change? Markram’s plan is to integrate the data from tens of thousands of brain research papers published every year, to take what is known about every section and every function of the brain and use it to create an unfathomably complex model “from the genetic level, the molecular level, the neurons and synapses, how microcircuits are formed, macrocircuits, mesocircuits, brain areas — until we get to understand how to link these levels, all the way up to behavior and cognition.”
Even today’s most powerful supercomputers can’t approach that level of simulation. But by 2020, they might. Ultimately, the code developed for what Markram calls the Human Brain Project would be available to any researcher. It would allow them, for instance, to flood the virtual brain with programmed versions of experimental drugs or disrupt brain functions and observe what happens. This could be a brain research motherlode, one that boosters insist could dramatically improve the ability to map and conceivably, cure brain diseases, all while reducing the need to experiment on animals’ brains.
Now who could have a problem with that? Turns out plenty of scientists do. They say that Markram’s guilty of extremely wishful thinking, that his approach to simulation just won’t work or would be no easier to understand than the brain itself. And they fear that if the EU pumps a billion euros into the project, all neuroscience would be subsumed by Markram’s vision.
Still, he contends that if not him, someone else will develop a virtual brain that works like the real thing. “Simulation-based research is an inevitability,” he told a meeting of scientists earlier this year. “It has happened already in many areas of science. And it is going to happen in life science.”
The brain that wouldn’t die
Okay, now let’s take the brain in a different direction, albeit one that might seem equally fanciful. A few weeks ago, at the Global Future 2045 International Conference in Moscow, a young Russian media mogul named Dmitry Itskov shared his dream of brains unleashed. Phase one of his project, which he calls Avatar, would involve controlling robots with human brains. That’s not so far-fetched. DARPA, the Pentagon research agency, is ramping up its own project, also called Avatar, in which soldiers would control with their brains a mechanical surrogate.
But then Itskov takes the fast train to fantasyland. Phase two would involve “transplanting” a human brain into a synthetic body. He thinks that’s doable within 10 years. And 30 years from now, he believes it will be possible to develop hologram-type bodies that can host an artificial brain rather than a physical one–now he’s speaking Markram’s language. That, claims Itskov, “would be leading down the road to immortality.”
Itskov acknowledges that this can sound like crazy talk. But, he notes, they said the same thing about the Internet.
Here’s more of the latest research on how our brain does what it does:
- Quiet down in there: New Scientist writer Sally Adee says a “thinking cap” is quite plausible after being wired with electrodes that stimulated her brain to induce “flow states.” The stimulation, she says, made her brain shut out all distractions.
- Even Google Maps hasn’t gone there: A state-of-the-art imaging scanner, developed by Siemens, has started mapping the brain in great detail by tracking the passage of water molecules through nerve fibers.
- In the weed: A Canadian researcher studying how marijuana affects memory believes that brain cells other than neurons help determine what we remember. It’s long been thought that neurons do all the heavy lifting in storing memories, but scientist Xia Zhang says his research shows astroglial cells also are involved.
- You’re so in my head: Scientists at the University of Technology in Sydney say that harmonious couples can actually be “on the same wavelength.”
- The secret to winning your March Madness bracket: In short, go with your heart. A study at Columbia Business School found that people who were more likely to trust their feelings were also more likely to accurately predict the outcome of events.
Video bonus: The video’s a few years old, but this TED talk by Henry Markram gives you a good idea of what’s going on inside his brain about brains.