Main | "3 LAWS UNSAFE" FLYERS FOR "I, ROBOT" »
July 15, 2004
PRESS RELEASE: SINGULARITY INSTITUTE RELEASES NEW WEBSITE ON AI ETHICS COINCIDING WITH “I, ROBOT” FILM
![Singularity Institute for Artificial Intelligence [SIAI] logo](/images/siai.jpg)
Atlanta, GA – In anticipation of 20th Century Fox’s July 16th release of I, Robot, the Singularity Institute announces “3 laws unsafe” (www.asimovlaws.com). “3 laws unsafe” explores the problems presented by Isaac Asimov’s Three Laws of Robotics, the principles intended for ensuring that robots help, but never harm, humans. The Three Laws are widely known and are often taken seriously as reasonable solutions for guiding future AI. But are they truly reasonable? “3 laws unsafe” addresses this question.
Asimov’s Three Laws of Robotics:
- A robot may not injure a human being or, through inaction, allow a human being to come to harm.
- A robot must obey orders given to it by human beings except where such orders would conflict with the First Law.
- A robot must protect its own existence as long as such protection does not conflict with the First or the Second Law.
Tyler Emerson, Executive Director of the Singularity Institute: “The release of I, Robot is a wonderful chance to engage more people about the perils and promise of strong AI research. The constraints portrayed in I, Robot appear extremely dangerous and quite lacking as an approach to moral AI. The Singularity Institute’s detailed approach, by contrast, utilizes advanced technical research for creating Friendly AI, a mind that is humane in nature.”
“3 laws unsafe” will include articles by several authors, weekly poll questions, a blog for announcements and commentary related to I, Robot and the Three Laws, a free newsletter subscription, and a reading list with books on relevant topics such as the future of AI, accelerating change, cognitive science and nanotechnology.
The Singularity Institute’s Advocacy Director, Michael Anissimov: “It is essential that more considerate thinkers get involved in dialogues of AI ethics and strategy. Although AI as a discipline has a dubious history of false starts, the accelerating growth of computing power and brain science knowledge will very likely result in its creation at some point. In the past few years, technologists such as Ray Kurzweil and Bill Joy have been informing the public about this critical issue; but more awareness is now needed.”
The Singularity Institute for Artificial Intelligence (SIAI) was founded in 2000 for the pursuit of ethically enhanced intelligence by creating humane AI. SIAI believes the ethical and significant enhancement of intelligence will help solve contemporary problems, such as disease and illness, poverty and hunger, more readily than other philanthropic causes. SIAI is a tax-exempt non-profit organization with branches in Canada and the United States.
For more information, please visit www.singinst.org.
Posted by SIAI at July 15, 2004 05:46 AM
Comments
May human beings create an intelligence that is morally advantageous to our future survival and comfort? Humans by our nature had to be aggressive to overcome our lack of natural weapons. Will we be able to create intelligence from scratch that will be morally superior to ourselves? Only time will tell and I am more than willing to argue both sides of the arguement.
Posted by: Richard Barniak at July 25, 2004 02:44 PM
Actually, in the end of 20th century, the 3 laws of robotics has been extended to following :
The Meta-Law
A robot may not act unless its actions are subject to the Laws of Robotics
Law Zero
A robot may not injure humanity, or, through inaction, allow humanity to come to harm
Law One
A robot may not injure a human being, or, through inaction, allow a human being to come to harm, unless this would violate a higher-order Law
Law Two
A robot must obey orders given it by human beings, except where such orders would conflict with a higher-order Law
A robot must obey orders given it by superordinate robots, except where such orders would conflict with a higher-order Law
Law Three
A robot must protect the existence of a superordinate robot as long as such protection does not conflict with a higher-order Law
A robot must protect its own existence as long as such protection does not conflict with a higher-order Law
Law Four
A robot must perform the duties for which it has been programmed, except where that would conflict with a higher-order law
The Procreation Law
A robot may not take any part in the design or manufacture of a robot unless the new robot’s actions are subject to the Laws of Robotics
Posted by: lawrence at July 31, 2004 10:19 AM
The laws are well and good, but it does not seem to me that they are so protected that a glitch in a machine or a hacking job may not remove or alter the rules. The more complicated a machine gets the more likely they are to develop unexpected difficulties. Look at the amount of difficulties that are inherent in windows operating systems. The software will develop errors simply by turning the machine on and off. What kind of errors may we expect in a machine that emulates and maybe even surpasses our human abilities? While the laws may work on 99% of the robots that are manufactured with them, what about the 1% or smaller that may have difficulties? I believe I am looking at this realistically when auto manufacturers have a planned obsolescence built into vehicles so you will need to purchase a new vehicle in five years or so. Hopefully, the manufacturers of robots will not follow this proven path of production.
Posted by: Richard Barniak at August 6, 2004 10:58 PM
A few questions regarding “Law Zero” posted above: where does one draw the line between a nuclear launch and, say, a freon powered refigerator? Assuming that robots would have a well designed AI, they would certainly be capable of making inferences such as “this man is releasing noxes in the atmosphere; eventually, noxes will lead to pollution, etc; therefore, this man is injuring humanity; according to law zero, I may not allow this”.
Also, let us consider a case similar to the famous Cuban missile crisis. Let us further consider that in this (similar, hypothetical) case, the US have lots of robots operating their weapons, while Russians are ignorant of the technology, but well aware of the seven laws of American robots. Should Russians decide to launch, what would the robots’ reaction be? Of course, they would be ordered to answer back; but launching a nuclear weapon, even in this situation, would still injure humanity (and would certainly not stop the harm already done by the first launch). Therefore, Law One would stop these robots from firing back; even more, suppose the robots are not the ones responsible for firing; they would still be bound to stop anyone from doing so, and attack their very creators in the process. Who would want any of this, I wonder?
To conclude, I think that the very interpretation of these laws relies heavily on the AI entity’s understanding of the concepts involved, which may change in time (for a genuine, learning AI). And “hard-wiring” all possible cases seems at the very least a colossal task.
Posted by: Daniel at August 23, 2004 05:53 PM