From brains to computers

By Amy Adams

So let’s say you want to make a piece of electronics that works just like the brain. Where would you start?

That’s the question neuroscientist Bill Newsome, PhD, director of the Stanford Neurosciences Institute, posed in a recent talk to a Worldview Stanford class on decision-making.

I thought the idea was so intriguing I wrote a series of stories about what it would take to reverse engineer the brain, and how close we are to succeeding at each. We’re still a ways from computers that mimic our own agile noggins, but a number of people are making progress in everything from figuring out where the brain’s wiring goes to creating computers that can learn.

These are the steps Newsome outlined to take us from our own grey goo to electronics with human-like capacities:

  1. Map the connections: Neuroscientists Karl Deisseroth, MD, PhD, and Brian Wandell, PhD, are mapping where the brain’s 100 billion neurons go.
  2. Monitor the signals: Biologist Mark Schnitzer, PhD, and bioengineer Michael Lin, MD, PhD, have created ways of watching signals in real time as they fire throughout the brain
  3. Manipulate the system: Neuroscientists Karl Deisseroth, MD, PhD, and Amit Etkin, MD, PhD, are working on techniques to manipulate the way the brain works and watch what happens.
  4. Develop a theory: Not only do we not know how the brain works, we don’t even really have a theory. Applied physicist Surya Ganguli, PhD, is working to change that.
  5. Digitize the circuits: If you want to turn the brain into electronics you need some wiring that mimics the brain. Bioengineer Kwabena Boahen has made just such a chip.
  6. Teach electronics to interact: Engineer Fei-Fei Li, PhD, has taught a computer to recognize images with almost human-like precision. This kind of ability will be needed by electronics of the future like self-driving cars or smarter robots.