Building the Better Brain: Ray Kurzweil on why reverse engineering the human mind is just to be expected
How will you know when an intelligent machine has reached consciousness?
For heralded inventor and futurist Ray Kurzweil, it will be when the machine can make him believe that it’s mad.
“If an entity really is convincing in its emotional responses, it has all the subtlety associated with this behavior, really reflecting the subjective state that it appears to be in – unlike say a video character in a video game that says ‘I am angry at you’ but you don’t really believe it yet – when we do believe it, when it really is convincing, with an emphasis on ‘convincing,’ then I will accept it,” Kurzweil told a near-capacity crowd gathered last night at the Kentucky Center for the Arts’ Bomhard Theater.
Kurzweil offered his personal definition of “consciousness” in response to an audience member’s question about the difference between a “brain” and a “mind,” on-point given that the best-selling author was in town to promote his new book, “How to Create a Mind: The Secret of Human Thought Revealed.”
“How to Create a Mind” details Kurzweil’s theories on how the human brain functions and how reversed-engineered technology can mimic and dramatically expand those capacities.
“A ‘mind’ is a ‘brain’ that’s consciousness has free will and identity,” Kurzweil initially replied to the audience member’s question.
Then, as super-intelligent people tend to do (Kurzweil holds 19 honorary degrees and, at age 17, famously showed off a music-composing computer he built from scrap on the TV game show “I’ve Got a Secret”), he went on casually cite various philosophical, scientific and hokum (see: Quantum Consciousness) standards for identifying “consciousness” – all of which ultimately boil down to a subjective “leap of faith” by the entity interacting with another “mind,” whether biological or mechanical.
“You have to have a leap of faith when it comes to who or what is conscious, or we really couldn’t function in this world,” Kurzweil said. “There’s no objective test for subjectivity.
“There’s a conceptual gap between the two … I maintain there is no machine you could build, you slide an entity in and a green light goes on and says, ‘Ok this one is conscious and this one isn’t,’ without building in some philosophical assumptions.”
Most of the evening’s conversation was not quite so philosophical, instead tending toward hard-wired topics such as the hierarchical cognitive structure of the brain’s neocortex via what Kurzweil has dubbed “pattern recognizers,” and the potential for 3D transistor technology to create a sixth paradigm for Kurzweil’s own Law of Accelerating Returns that by the 2030s will result in everyone having blood-cell-sized computers in their brains to connect them directly to the Cloud.
Stuff like that.
(A gracious woman sitting next to me in the Bomhard, upon seeing my reporter’s notepad, suggested that I would be fortunate to understand enough of what was about to be said so that I could write it down.)
Kurzweil and interviewer Jim Fleming also touched on topics ranging from the opportunistic nature of successful innovation to the near-term implications of nanotechnology even before we reach the technological singularity, where human consciousness sort of lives forever in a cyber-super-intelligence, a concept Kurzweil famously (and controversially) has espoused in his books such as “The Age of Spiritual Machines and Fantastic Voyage: Live Long Enough to Live Forever.”
The event, part of the University of Louisville’s Kentucky Author Forum, was taped for later airing on KET’s nationally syndicated Great Conversations series, as well as for audio broadcast locally by WFPL.
In the meantime, check out this video of Kurzweil’s presentation at this year’s Learning Without Frontiers Conference in London. Even if you don’t want to wind up as a ghost in the machine, he is a genuinely fascinating speaker.
Last night, Kurzweil attributed his lifelong fascination with science to his grandfather, who on his first trip back to Europe after fleeing fascism had the chance to handle documents created by Leonardo. These documents had a sense of the “sacred,” Kurzweil said, but a sacredness that arose from not the divine, but human potential.
“It was personalized,” he said about his grandfather’s encouragement. “You, Ray, can change the world.”
After making his own rocket kits and puppet theaters as a child – “This was an era when you let an 8-year-old roam the neighborhood and bring home spare things,” he joked – Kurzweil went on to a groundbreaking career in optical character and computer speech recognition, creating the first commercially marketed speech recognition software.
And then there is his often controversial work as a futurist, in which he consistently asserts that humans will use technology to radically expand and redefine ourselves.
Kurzweil doesn’t like the term transhumanism, even though he is often cited among the movement’s leading thinkers. To him, it’s simply human nature to use technology to do things we otherwise could not. “We are the species that aspires to expand beyond our limitations … We are the species that changes who we are,” he said.
In his latest book, “How to Create a Mind,” Kurzweil draws on his research in cognitive and language patterns to describe the function of the neocortex, the nifty part of mammalian brain that holds our higher functions and let us come out ahead after that Cretaceous Extinction business.
Kurzweil suggests we have about 300 million “pattern recognizers” in our neocortex that store and organize data in a complex hierarchy. Once those 300 million pattern recognizers are full, the brain must overwrite old (often redundant) data to learn new things.
It’s why children pick up new languages more quickly than adults, and why Einstein was a lousy violinist – there’s only so much storage and computational brain power to go around.
Within 30 years or so, Kurzweil believes nanotechnology will have advanced to where people will simply have computers embedded in their brains that connect to the Cloud (or whatever it’s called then) to add “pattern recognizers” when needed. To Kurzweil, it’s not at all unlike his own smartphone, which he repeatedly called a “brain extender” during the 90-mintue conversation last night.
The trick will be reverse-engineering the hierarchical functions of neocortex into the supporting technology, and Kurzweil believes science is well on the way to cracking that nut. He noted the Watson super-computing project that “read” 200 million Wikipedia pages and then went on to win big at Jeopardy!
In the more immediate future, he said, expect nanotechnology to soon correct our white blood cells’ confusion about cancer being OK.
Kurzweil stands behind the science of his predictions, saying that he’s been right about 86 percent of the time, not including near-misses like self-driving cars. (Google is making headway, but they are not yet broadly available.)
An audience member last night asked Kurzweil if he thinks a “noise” backlash will halt the march of all this nano-progress (of course, Nate Silver came up), but the futurist replied that the Luddites were wrong, and so are those who think technology will not continue to accelerate more and more rapidly.
Kurzweil did add that sometimes innovation needs a little help from good timing – other social networks were fretting over photo uploads before Facebook rode broadband and cute baby pics to world domination – but ultimately the trend lines don’t lie.
Not surprisingly, middle-schoolers, with all those empty pattern recognizers, have proven to be the most receptive audience to Kurzweil’s predictions, he said. “They say, ‘That’s so true – things were so different when I was 8.’”
While he was in town yesterday, Kurzweil also spoke to a group of students at U of L’s School of Business.