Revisiting Asimov's Three Laws of Robotics

by AMNH on

News Posts

The term artificial intelligence was coined by computer scientist John McCarthy in 1956.
Three humanoid robot prototypes are displayed standing side-by-side.
Humanoid robot prototypes, like these developed by Honda in the 1980s, were developed to help people with everyday tasks.
Morio/ Wikimedia Commons

Nearly 15 years earlier, Isaac Asimov published a short story that hinted at how complex our interactions with “intelligent” technology might be—and inspired some of the biggest thinkers in the fields of robotics and A.I.

The 1942 story, Runaround, introduced Asimov’s Three Laws of Robotics and the unintended consequences that could arise if these rules came into conflict with one another. The code forbids robots from hurting humans and mandates that they obey human commands. The story also highlighted the responsibility of those who create technologies to anticipate and safeguard against potentially threatening events. When the story was published along with eight others in the seminal 1950 sci-fi collection I, Robot, it introduced the Three Laws of Robotics to a broader audience.


Book cover with text reading Asimov, the Robot Series, I, Robot, featuring a robot standing with arm raised.
Isaac Asimov’s short story collection I, Robot, introduced a wide readership to the Three Laws of Robotics when it was published in 1950.
RA.AZ/ Flickr

In the last three decades, A.I. development has accelerated at an amazing pace—and gotten so entwined in our daily lives that we now start our days by asking Siri about the weather or count on Alexa to order our groceries.

Just as an example of how far and how fast A.I. has come in three decades: In 1997, IBM created the chess-playing Deep Blue computer, which was able to “recall” 200 million chess positions per second and beat then-reigning world chess champion Garry Kasparov. In late 2017, Google’s DeepMind initiative demonstrated that its AlphaGoZero program learned to play the Chinese board game Go from “scratch”—with no data from human games. And AlphaGoZero played at a higher level than all previous AlphaGo programs, one of which had defeated the Go world champion earlier that year. 

So is it time that Asimov’s Three Laws got an update? In 2017, the Asilomar Conference proposed 23 principles to guide the future of A.I. development and address issues including human control, safety, and the avoidance of an A.I. arms race.

Some of these topics will be up for discussion at the 2018 Isaac Asimov Memorial Debate as Neil DeGrasse Tyson, Frederick P. Rose Director of the Hayden Planetarium, explores the benefits and consequences of A.I. with Google Senior Vice President John Giannandrea, co-founder of iRobot Corporation Helen Greiner, IBM Watson Chief Architect Ruchir Puri, MIT professor Max Tegmark, and University of Michigan professor Michael Wellman. 

The 2018 Isaac Asimov Memorial Debate, which took place on February 13, 2018, is available as a podcast and as a video.