Robots: are we planting the seeds of our next evolutionary step?
August 22 2008 / by futuretalk
Category: Technology Year: General Rating: 9 Hot
By Dick Pelletier
While some scientists see danger with tomorrow’s super-intelligent robots, others say not to worry – our destiny is secure.
For years, entrepreneurs have been trying to create robots to perform life’s physical drudgeries. Building mechanical bodies has been easy, but creating artificial minds to control those bodies has been frustrating.
After countless commercial failures, things are beginning to change. Computer power has increased to 1,000 million impressions per second (MIPS), which provides enough thinking ability for today’s robots to become financially viable.
Lawrence Livermore National Labs robots will soon handle nuclear and hazardous waste. NASA Rovers explore Mars and send pictures back to us.
Sales of “Roomba” and “Karcher” robot vacuum cleaners are beating expectations. Thousands of Sony AIBO robot pets have sold for more than $1,000 each. Robots are now becoming main stream.
Research Professor Hans Moravec at Carnegie Mellon University says future robot development can be divided into four generations:
2010 – First generation robot will: pick up clutter; store; retrieve and deliver; take inventory; guard homes; open doors; mow lawns; and play games. Compare its 5,000 MIPS brain to a lizard.
2020 – Second generation robot will think before it acts and respond to dog-like training. “Good robot,” “bad robot.” Compare its 300,000 MIPS brain to a mouse.
2030 – Third generation robot will understand human moods. It knows if you are happy, angry, in a hurry, or tired. It handles limited conversations. Compare its 10 million MIPS brain to a monkey.
2040 – Fourth generation robot will simulate the world with powerful intelligence. Special sensors allow it to see through walls. It seamlessly interacts with humans. Compare its 300 million MIPS brain to us.
Fourth generation robots can also make improved copies of themselves, which could quickly outpace human evolution. This causes concern for some.
In a recent “Focus” magazine article, acclaimed scientist Stephen Hawking warned, computers are advancing faster than humans; if we don’t make changes, they could take over our world. Bill Joy agrees. In his manifesto “The Future Doesn’t Need Us,” the Sun Microsystems scientist expressed alarm that eventually our silicon creations might replace us.
But AI author Ray Kurzweil offers a different view. “As computers exceed the information processing capability of the human mind, the boundary between human and machine becomes increasingly blurred.” Kurzweil asserts that robots are a human creation, and tomorrow’s super-intelligent versions could simply merge with us and become the next step in evolution – The Human-Technology Civilization.
Moravec considers these future machines “mind children,” built in our image and likeness – human, but in more potent form.
Will these events happen? Futurists believe by mid-century, we could be swapping frail biological bodies for powerful, immortal silicon versions. Danny Hillis, founder of Thinking Machines Corp says “I’m as fond of my body as anyone, but if I can be 200 with a body of silicon, I’ll take it.” Comments welcome.
Comment Thread (2 Responses)
-
“As computers exceed the information processing capability of the human mind, the boundary between human and machine becomes increasingly blurred.”
I know that Kurzweil often expresses this view, but he still seems to back the Strong AI singularity. Still it’s good to hear that even he leaves the door open to a merge, though doesn’t lend a ton of credence to the Strong IA possibility in which some or many humans get vastly smarter by 2020, 2030, 2040.
I think we’ve already begun to blur with machines and that that began with the earliest symbology – technology is part and parcel of what we are – and thus robots will be viewed more as extension unless brains that process memes and temes independent of humans, as Susan Blackmore has proposed, can escape the human brain web. Iow, strong-enough AI developed without links to human emotional centers, which may or may not happen depending on how smart humans really are en mass and whether human networks will permit such intelligence to exist independently.
Posted by: Alvis Brigis August 22, 2008
Vote for this comment - Recommend -
I think before our human-machine merge can gather much steam, we must first 1) unravel the mysteries of human consciousness, which could happen by as early as 2020 and 2) program cognitive human abilities into robots enabling them to relate on a human level (emotions and all) to humans. This could happen by as early as 2030 and it will teach us all we need to know to eventually replace all our biology with non-biological nanomaterials giving us a manner of invincibility.
Comments welcome.
Posted by: futuretalk August 22, 2008
Vote for this comment - Recommend