Knowledge is power

It is common these days to hear in the news about a new robot that is about to change our lives or some scientist who claims that robots will soon be doing all the manual labor for us. Scientists are busy trying to create useful robots; these robots will perform important tasks such as caring for the elderly and going into places that humans can’t safely go such as deep underwater and underground. Mobile robots are already essential tools for planetary exploration; NASA has had lots of success using robot rovers to explore the Martian terrain and several countries are currently preparing robotic missions for Mars and the Moon.

But machines are becoming much smarter every year. There is lots of evidence to suggest that computers will eventually outsmart humans in the not so distant future. In some cases, artificial intelligence systems already outperform humans. Let’s look at a few of those.

Chess playing: IBM’s Deep Blue supercomputer was probably the first machine to defeat man in a game that many previously thought could not be played well by a machine lacking human intuition. On May 1997, Deep Blue defeated world chess champion Garry Kasparov to alert us on the coming of machines.

Checkers solved: 10 years after Deep Blue, scientists at the University of Alberta, Canada, announced that they had solved the game of checkers after more than 20 years of continuous computation. Solving a game means that no matter what the first move is by either player, the computer knows how not to lose. When it comes to checkers, computers know the strategy that they have to play such that the game always ends in a draw. Checkers is the largest game to date that has been solved. Remarkably, checkers has more than 500 billion billion (that’s two billion and not a typo) configurations making the computer’s achievement even more important.

Polaris poker playing AI: The same group from Alberta also developed the poker playing artificial intelligence program Polaris which recently played 2000 hands of Texas hold’em against two professional human players. Ali Eslami and Phil Laak played against Polaris in a 2-day tournament for the first ever man-machine poker championship. Ali and Phil managed to defeat Polaris but only by a small margin giving scientists hope that with a bit more tweaking their software will soon match Deep Blue’s performance in outperforming humans in game playing.

Robot soccer: The Robocup organization announced years ago that their ultimate goal is to develop a team of robots that is capable of defeating the human world champion soccer team by 2050. Soccer was first proposed as a good problem for AI to tackle back in 1992 in a paper written by Alan Mackworth. Since the late 90s, the Robocup organization has held an annual robot soccer competition to test the limits of artificial intelligence and robotics in the dynamic, competitive and cooperative world of soccer. In the last 15 years, Robocup has evolved from a tournament hosting slow and rather stupid robots trying to play soccer to one full of intelligent machines working together to score and defend. The humanoid league that was introduced a few years ago is also gaining lots of momentum with dozens of teams participating every year. It is conceivable that Robocup may achieve its ultimate goal by 2050 if not earlier.

Service robots: There are millions of service robots in use today. The robotic vacuum cleaner Roomba is the most popular and successful commercial robot every sold. Its creator, iRobot, has since expanded its line of domestic robots to included carpet and pool cleaning. The company also develops a line of robots used by law enforcement and the military saving lives in IRAQ and other war zones. General purpose, domestic robots are still the stuff of dreams but highly specialized ones have gained much momentum in recent years. As consumers become more comfortable with the idea that an automaton can do useful work and as prices drop, we are going to see more companies offering consumer robots to do all those jobs that humans hate doing on a daily basis.

The robots are coming and there is little that we can do to stop them. The sooner that we get comfortable with the idea that in the near future we will have to accept in our society artificial intelligence agents, the easier it will be for us to transition to our new role on Earth; that of being the creators of new lifeforms.