The blurb on the back of my copy of Asimov's The Complete Robot includes

Asimov's Three Laws of Robotics were programmed into computers thirty years ago at the Massachusetts Institute of Technology - with surprising results.

The book was published in 1995, so thirty years ago would have been mid 60s. Does anyone know what happened or what the surprising results were?

  • 8
    That sounds like advertising blurb to me - Asimovs laws are incredibly hard to translate into actual code, and would be beyond the systems and languages we have today, as they involve a lot of understanding of things that we cant just code into a neural net yet.
    – Moo
    Commented Oct 18, 2015 at 17:22
  • 2
    Asimov's laws are invented to better serve the literature than to be logically perfect or applicable in the real world. So most probably the result was a complete failure to achieve the main idea of the laws - to guarantee the safety of humans around robots.
    – vap78
    Commented Oct 18, 2015 at 18:48
  • @Moo That's not true. There actually were several experiments and translating the rules is rather easy - if you are not working with actual robots and humans. However, the first experiment I could find was in after 2000, not in the sixties. But it's definitely possible.
    – Sulthan
    Commented Oct 18, 2015 at 19:27
  • 3
    Asimov mentioned (in some compilation, I don't remember which one) that he came up with the rules as an extension of commonsense industrial safety thinking. Machinery (he posited) is designed to be 1) safe to operate, 2) must accomplish its intended purpose (but fail to operate if needed for safety), and 3) try not not to damage or wear itself during its operation (but suffer damage rather than endangering safety, plus some equipment needs to wear heavily to accomplish its purpose). In that sense, the laws have been applied for centuries. Commented Oct 20, 2015 at 4:02
  • 1
    There was a recent experiment in which a computer was programmed with such "rules" in order to determine whether to warn particular people about a hazard. The machine actually did not warn everyone, and thus caused some people harm. It was written up in an issue of "New Scientist" in the last 8 months. Of course, Rudy Rucker wrote a story in which a robot points out that the Asimov laws put humans first and self-preservation of the robot second at best. It responds "Forget THAT! no way!" Commented Nov 7, 2015 at 3:31


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.