We've posted more stories on robot ethics over the years than I can count. The general public and lawmakers still seem ignorant of the issues and even many roboticists still seem unclear on the import of autonomous robots that can make the decision of when to kill and who to kill without a human in the loop at all. But each day we come closer to having fully autonomous war robots. Some researchers, like Ronald Arkin, believe we can create robots that fight only for us and kill only in an ethical fashion. Other researchers, like Noel Sharkey, have warned that we shouldn't build autonomous weapons and that, if we do, they will eventually be copied and turned on us as well. A good comparison of Arkin's and Sharkey's views can be found in Part 1 and Part 2 of the Robots podcast on Robot Ethics. The latest development in this area is the publication of a 50-page report by Human Rights Watch called Losing Humanity: The Case against Killer Robots (PDF format). From the report:
Based on the threats fully autonomous weapons would pose to civilians, Human Rights Watch and IHRC make the following recommendations, which are expanded on at the end of this report:
- Prohibit the development, production, and use of fully autonomous weapons through an international legally binding instrument.
- Adopt national laws and policies to prohibit the development, production, and use of fully autonomous weapons.
- Commence reviews of technologies and components that could lead to fully autonomous weapons. These reviews should take place at the very beginning of the development process and continue throughout the development and testing phases.
Noel Sharkey was a technical reviewer on their report and provided input. Even if it's too late to stop the development of fully autonomous robots or Apocalyptic AI, it's well worth reading this report, which covers a lot of interesting points. If not an outright ban, it's likely we'll at least seen changes to existing laws as well as new ethics requirements researchers. While we have yet to reach the point of fully autonomous killer robots, several robots are quite close, such as the Northrop Grumman X-47B, pictured above, and the Samsung semi-autonomous Techwin SGR-A1 border guard which can fire on and kill humans. If you're not up for reading the full report, read on to see the short video released by the Human Rights Initiative to summarize the content.
CC licensed image of X47B from flickr user US DoD