The Rise of the Artifical-Intelligence Economy

Adam Ozimek -- blogger at Modeled Behavior and associate at Econsult Corporation

As a child I used to read my grandfather's Popular Science and Popular Mechanics magazines. The constant promise and inevitable disappointment of amazing technologies that mostly never materialized (a problem likely exacerbated by my focus on the amazing and outlandish ones) made me skeptical of futurist predictions. It is somewhat strange then, that I now commonly find myself a proponent of futurist visions equally as grand as those that once made me a cynic. But I'm not alone in seeing the near future as a quickly changing technological landscape. In their recent book Race Against The Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy, MIT's Erik Brynjolfsson and Andrew McAfee* offer a similarly sweeping view of how technology is, and will be, shaping our future

They present two convincing cases of technology changing quicker than we would have thought. First is the driverless car. This is something that Popular Science has been promising since the 1940s, and I remember reading about 20 years ago. But despite the optimistic predictions of yesterdays futurists, a car without a driver was always a farther off than it seemed. As recent as 2004, in fact, economists Frank Levy and Richard Murnane argued that the kind of pattern recognition that driverless cars required was impossible:

The... truck driver is processing a constant stream of [visual, aural, and tactile] information from his environment. ... To program this behavior we could begin with a video camera and other sensors to capture the sensory input. But executing a left turn against oncoming traffic involves so many factors that it is hard to imagine discovering the set of rules that can replicate a driver's behavior. ...

In that same year DARPA held their first Grand Challenge, which asked competing teams to build a driverless car that can make it across 150 miles of dessert. Confirming Levy and Murnane's pessimism, the longest any car made it was 8 miles, and this took several hours.

Yet despite how difficult this challenge seemed just a few years ago, Google has made astounding headway in building a functioning driverless car. As the video below shows the current capabilities are already very impressive. So much so that the state of Nevada recently became the first state to pass regulations allowing autonomous cars.

Skeptics cite our deep aversion to handing over control to a computer as an impediment to the driverless car. But it need not be the case that the first time you hand control to a robot it will have you barreling down the interstate at 70 miles-per-hour. Autonomous driving might first be used for slow moving, stop-and-go traffic. You can see a precursor to this in cars that are already parking themselves. We can ease our way into comfort with it. We should have little doubt: driverless cars are in our future.

Brynjolfsson and McAfee also cite IBM's Jeopardy! winning supercomputer Watson as further technological proof that we are on the cusp of changing our world. Watson shows more of the impressive pattern recognition seen in driverless cars, but also demonstrates complex language skills that were once thought beyond the province of computers. Supercomputers like Watson will drastically change medicine and other knowledge fields, and in fact IBM and Memorial Sloan-Kettering Cancer Center are already working on teaching Watson to aid in diagnosis and suggesting treatments for cancer.

Others are joining in on the chorus and heralding the new age of Artificial Intelligence. In a new piece in The American Interest, Tyler Cowen argues that AI is one of the three reasons the United States is becoming an exporting powerhouse:

.... artificial intelligence and computing power are the future, or even the present, for much of manufacturing. It's not just the robots; look at the hundreds of computers and software-driven devices embedded in a new car. Factory floors these days are nearly empty of people because software-driven machines are doing most of the work...

...The next steps in the artificial intelligence revolution, as manifested most publicly through systems like Deep Blue, Watson and Siri, will revolutionize production in one sector after another. Computing power solves more problems each year, including manufacturing problems.

If we accept that this increase in technological growth is occurring, it begs the question of "why now?" The authors argue that this is happening now because of Moore's Law, and variations of it, which predicts that specific technologies will double in performance at regular intervals. Exponential growth like this means that we will experience a period of what looks like slow linear growth followed by a fast acceleration of growth that "confounds expectations and intuitions." They believe we are entering the period of fast acceleration.

If Brynjolfsson and McAfee are right, we should all be futurists now. Technologies that recently seemed implausible will soon become reality. A long time skeptic, I find myself thinking more and more like this today. It certainly seems like a weekly basis that a Youtube video emerges of a technology that "confounds expectations an intuitions". From the creepily humanoid robotics of Petman, to somehow menacing cooperation of a swarm of nano quadrotors, things seem a little more futuristic lately.


*An earlier version of this article incorrectly referred to Andrew McAfee as Eric McAfee.