While in Berlin, Germany last week with Ron Arkin and Colin Allen, the IEET published the question, "Do we need a law making it illegal for computers and robots to deceive or be dishonest?" The question had been stimulated by recent articles about research performed by Ron and research engineer Alan Wagner. Ron was particularly pleased that this research had gotten people to ask questions such as this. Stimulating reflection on serious ethical concerns has always been one of his goals.
However, our conversations went in a somewhat different direction. Is the publicity creating the impression that the relatively low level mechanisms Arkin and Wagner introduced into their experiment are the equivalent of higher level cognitive ability? In other words, are we feeding a false impression that robots are much more sophisticated than they are, or are likely to be in the foreseeable future?
Ron pointed out that the actual research and the press release that accompanied it was responsible, and as we all know the press can distort scientific findings for its own purposes.
Here are some addition links for those interested in this subject.
Click here to link to the research paper titled, Acting Deceptively: Providing Robots with the Capacity for Deception
Hyperlink to the original Press Release.
Article at NewScientist titled, Deceptive robots hint at machine self-awareness.
Vote on the question at Polldaddy and view the results.
While roughly 60% favor outlawing or restricting deceptive robots, b ut only half of these thought it enforceable.