With nanotech, expanding the brain to the cloud

Author, inventor, entrepreneur, and futurist Ray Kurzweil has accurately predicted the rise of major technological innovations, from head-mounted displays such as Google Glass to natural language interfaces such as Siri.

Speaking Feb. 20 at the annual MIT Tech Conference, Kurzweil offered a vision of 2030, one with nanorobots bolstering the immune system and also connecting to external, cloud-based neocortal modules, or groups of neurons, to access far more knowledge than can fit in the brain.

“We have pretty good ideas of how this works,” said Kurzweil, who detailed this process in his 2012 book, How to Create a Mind. “There’s a scenario for every different type of disease … a tool for doing surgery at the microscopic level.”

Thinking exponentially, not linearly

The thought of nanorobots inside the body fighting disease and connecting to computers may seem far-fetched. But so did, at one time, the World Wide Web, the mobile phone, the 3-D printer, and the fully mapped human genome—all of which Kurzweil also foresaw.

Human intuition about the future has always been linear, said Kurzweil, a 1970 MIT graduate and current director of engineering at Google leading efforts to build artificial intelligence and natural language understanding. This dates back to the days of hunting game on the savannah, when humans could literally only see what was in front of them. However, technological innovation has grown exponentially since 1980, when U.S. Census data was tabulated by machine for the first time.

As an example, Kurzweil pointed to the Human Genome Project. It began in 1990; seven years later, it was only 1 percent complete. Critics, thinking in a linear way, said the project would take 700 years. Kurzweil, thinking exponentially, said it was half-finished. The project was declared complete in April 2003. When the project began, mapping an individual genome cost nearly $1 billion; today it can be done for under $1,000.

Nanorobots. Soon.

The emergence of nanorobots in the next 15 or so years will come about due to two related factors, Kurzweil said.

First, though the fifth paradigm of computing—the integrated circuit—is due to end by 2020, the sixth paradigm—3-D molecular computing—is already under development. Second, the rapid growth of computing power has brought with it extreme amounts of data, both of which drive artificial intelligence systems. The brain’s electrochemical signals are much slower than today’s electronic systems. The brain makes up the difference with its 3-D “circuitry,” Kurzweil said.

As that gap closes, Kurzweil foresees the rise of artificial intelligence-powered nanorobots fighting disease in the bloodstream and connecting to an external neocortex that he said can more than triple the number of neocortal modules to which an individual brain has access. It will have an even more dramatic effect on human evolution than the elongation of our foreheads, and the expansion of our neocortex, about 2 million years ago, he said.

“We’re going to do it again. Once we can expand our neocortex into the cloud, it won’t be limited by an enclosure [the skull],” Kurzweil said.

While some critics suggest that the rise of artificial intelligence will lead to a dystopian world, Kurzweil said the ongoing development of ethical rules for bioengineering will ensure that the perils of artificial intelligence depicted in science fiction can remain in check.

“I come out optimistic that we can pull through,” he said.