On several occasions during his decades as an IBM executive, former CEO Thomas J. Watson Jr. went out of his way to differentiate between human brains and computers. The goal of IBM, he said, was to make machines that could serve as tools for humans, not replace them. By taking on many of the more routine tasks that people were saddled with, he said, machines would free humans up to do more fulfilling creative work.
Today’s computers are “smarter” than those of the mid-20th century, when Watson was busy trying to calm people’s fears about the onslaught of electronic brains. But even with the achievements and potential of Watson, the Jeopardy-playing computer, IBM is pushing the borders of machine capability with the goal of augmenting human thought, not replacing human brains with machines.
Still it’s intriguing to compare the two, and for scientists, nature offers models for next-generation computer design. At the conference IBM Research Colloquia – Zurich last week, Prof. Karlheinz Meier of Heidelberg University spoke about neuromorphic computers during his presentation on Brain Inspired Computing. A key aspect of neuromorphic computing is understanding how the brain works and attempting to mimic some of its functions in silicon.
For me, the most intriguing moment during Meier’s presentation was when an audience member asked him a question that got to the heart of the man vs. machine debate. Essentially, he asked: Why copy nature? Why not aim to do better?
Question: “This may be a philosophical question. If you look at the properties of neurons and synapses and so on, do you see these wonderful devices or would it be possible to conceive that the brain evolved out of very simple cells, which represented a limited technology, and this was the best nature could do with what was available. If so, is it possible to invent other sorts of devices which could produce a better brain?
Meier: “Absolutely. That is a very valid question. We are today trying to copy what evolution has produced. It may not be the best thing. There may be more efficient solutions. That’s something we have to work with.”
One newsy element of the Zurich colloquium was the official launch of of a new $90 million Nanotechnology Center at IBM’s research laboratory in Zurich. The center is a joint venture of IBM and ETH Zurich, a premier European science and technology university. Meier said that the center, with it’s large collection of state-of-the-art nanotechnology tools, will be a valuable resource for scientists like him. “With this kind of facility and the tools and a rapid design cycle, we can try out all of these things,” he said.
In that building, where physics, biology, chemistry, and material science meet, the question of whether or not machines can be made that think may finally be resolved.