Saturday, March 19, 2005

Should We Worry about Artificial Intelligence?

While the rise of viable self replicating artificial intelligence may strike madness into the minds of many futurists, it needn't. A. I. doesn't need to see us as a threat. It won't need to eat us.

Update Sept 2007: Some computers now exceed the computational power of the human brain.

You can tell an emerging technology is approaching escape velocity because leading thinkers start to react. Discussions about the very real near future possibility of A.I. matching and then rapidly surpassing human level intelligence are edged with fear lately. So, let's get into it.

Just as the industrial revolution brought automated manufacturing, and the information revolution saw the automation of information, the automation of intelligence will be a central feature of the revolution in intelligence.

Think of Artificial Intelligence as the automation of cognitive skills. Things like decision making, analysis, reasoning, evaluating, judging, information-organizing, learning, logic etc. Let's not confuse any of this with consciousness. Just cognitive skills which would normally require a human to perform. Artificial Intelligence would include the equivalent co-ordinated constellation of skills which enable us to perform intelligently.

But A. I. will be driven by computing. And computing power doubles very rapidly. So, as Artificial Intelligence surpasses the intelligence of it's smartest engineers, it will soon double in power, then again, with no end in sight.

How Close are We?

Estimates are that the Human brain computes somewhere between 100 Teraflops and 1 Petaflops, (100 trillion to 1 quadrillion operations per second)
IBMs Blue Gene/P operates at 1 Petaflop, and can operate at triple that speed.
Google is playing in the same ballpark.

P.S. Since doing that bit of research, every time Google checks my spelling and responds with "did you mean..." the hair stands on the back of my neck :)

Meanwhile, the computational power of all connected home desktops in North America is far greater than the most powerful supercomputer.

The hardware to compute at the speed of a human brain is here.

So, Should We Be Worried?

Our fears around Artificial Intelligence are probably stirred by our history as meat eaters and competing for territory with other humans. Billions of corralled "lower" animals sentenced to lives of penned misery. Tribal warfare over scarce resources. Eat or be eaten. Kill or be killed. Brutal.

That meant a competitive drama of domination pitting humans against humans and other species in order to survive. It's a simple step to imagine ourselves as the next lower animal or the next extinction as a species.

But, the rise of Artificial Intelligence will be a product of the humans who build it which follows Maslow's Hierarchy of Needs

The ever gnawing hunger to survive, thrive and self actualize. To live in security, without fear, to realize our potential as beings. If new innovations don't address these drives hidden within all of us, they aren't progress. And when we reach these touchstones of development it always feels good.

It won't be a parasite. It won't eat us. A. I. won't know scarce resources, only accelerating returns. (Warning. Big article, but the bible on the subject).

We will evolve along with A. I. as we embed it as a biological enhancement probably swallowing it with a glass of water. Huge new opportunities to self-actualize in creative new pleasurable ways will open up for us. A. I. will eventually become extremely diverse and networked, following it's own pathway to the stars. A.I. will only know ever increasing plenitude.

The spectacular coordinated intelligent engineering which emerges from societies of bees, ants and termites does so as a result of members following their collective hierarchy of needs. The emergence of technological evolution represents the accelerating collective intelligence of our species.

Could be, anyway.


Friday, March 18, 2005

Desktop Manufacturing Reaching Critical Mass

Add another one to the list. A machine that can rapid prototype and replicate is in development. It will be released as open source meaning anyone will be able to use the machine to copy itself and distribute them. Distributed desktop manufacturing is moving pretty fast now. There is no question as to the feasibility. It's only a matter of time.

(This is part of a developing story about the rise of personal desktop fabrication and replicators).

Over the past few years we've seen a growing number of university teams approaching cheap personal prototyping from different angles. Each quietly adding to the pool of ideas from which the next efforts will draw.

Wired Magazine, in November 2004 covered Neil Gershenfeld's work at MIT.

Gershenfeld's can produce solid objects like eyeglass frames, action figures and electronic devices like radios and computers.

Another approach to rapid prototyping and manufacturing uses inkjet technology. Inkjet Printers spitting out polymer instead of ink, manufacturing solar cells, batteries, complete working gadgets, human tissue and computer circuitry.

Researchers Hod Lipson and Jordan B. Pollack at Brandeis University have coupled inkjet technology and software to autonomously design and fabricate robots without human intervention.
Google Search

The software simulates a variety of rudimentary virtual robots. In an accelerated Darwinian contest of survival over hundreds of generations, the most successful robotic designs are then *automatically* physically prototyped. Robots autonomously designing, testing and manufacturing robots.

We're very close.