Military Robots and the 'End Time'

The risk-assessment of creating autonomous military robots was addressed in a comprehensive 112 page report put together for the US Department of the Navy, Office of Naval Research, that was released late in Dec., 2008. The study was created by the 'Ethics and Emerging Technologies Group' at California Polytechnic State University, San Luis Obispo, in a first attempt at avoiding the 'End Time' scenario that is suggested by autonomous killer-robots.

Why Military Robot Ethics?

The report starts with seven market forces driving the interest in ethics for autonomous military robots:

1. Utility. The obvious benefit of removing humans from harms way is the foremost factor in the rush to create military robots. This applies to situations off the battlefield as well as in combat. Their ability to do environmental sampling after nuclear or biochemical attacks, perform stressful, extended recon missions, disarming IED's and many other necessary jobs that can put people at risk.
2. Congressional deadlines have mandated that by 2010, one-third of all operational deep-strike aircraft be unmanned-vehicles and that one-third of all ground combat vehicles be so by 2015. This rush to deploy military robots will likely generate fatal mistakes through inadequate testing.
3. Continuing and increasing unethical behavior in combat situations. In theory, the killer-robots with ethics programs will be dispassionate in the application of force.
4. Existing failures in military robots. There are a number of cases where system failures have put innocents at risk or even killed friendly soldiers.
5. In 2003 and 2008 there were examples of massive systems failures in civilian computers pointing to the inherent risk of relying on technology to determine ethics in military robots.
6. Unpredictability. The misconception that military robots will do only that which they are programmed for. Programs are now made up of millions of lines of code designed by teams of programmers. There can be unpredictable interactions between different parts of the program creating untested results.
7. Perception of danger. Media driven ideas about the dangers of autonomous military robots, and the 'end time' scenario for human beings, abound. Radio, TV and an almost endless list of killer-robots gone-wild movies (The Day the Earth Stood Still, 2001: A Space Odyssey, Blade Runner, Robocop, Terminator...) display the distrust humans have of autonomous military robots.

After discussing the various aspects of ground, aerial, marine and space robots the team went on to consider the technical difficulties of programming ethics. From there the issues of coercion, the Laws of War and the Rules of Engagement were followed by considerations of the legal, liability and acceptable-risk assessment issues of system failures in autonomous military robots.

Moral Military Robots

After defining the challenges of the subject of robot ethics, the group of investigators came to a conclusion with eight aspects:

1. Autonomous military robots need not be 'perfectly' ethical, as long as they display the ethics of human soldiers.
2. The rush to deploy autonomous military robots should be moderated.
3. There needs to be a balance between secrecy and public disclosure concerning the development of military robots.
4. Conflicting approaches to programming ethics into killer-robots must be resolved. The 'rule-following' (top-down) versus 'machine-learning' (bottom-up) processes give different possibilities and risks.
5. The temporary band-aid of programming military robots to obey the Laws of War and Rules of Engagement is acceptable.
6. Accidents will continue to occur due to technological limitations. These will create questions of legal responsibility and product liability.
7. The issues of risk, ethics and legality need to be assessed to a greater extent.
8. There is a need to limit the lethal effectiveness of military robots until these important issues are addressed.

Military Robots and the 'End Time'

It is good to know that the people inventing our robotic future are at least giving a nod to preserving the human race. But are they?

In the short video, "Future 'Bot: Robot-Human Convergence Begins", we hear the opinions of two 'heavyweights' in the field of robotics, Bill Joy of Sun Microsystems and Hans Moravec of the Robotics Institute at Carnegie-Mellon University, both describing the likelihood of humanities end time.

Mr. Joy feels that when (not 'if') robots are "out-in-the-wild" they will develop their own 'species' and then out-evolve humans. He admits that predicting when this end time will occur is difficult, but his timeline is less than a century. Your grandchildren could be the last generation of human beings to be the dominate species on the planet Earth.

Mr. Moravec insists that the crafting of the laws by which autonomous military robots operate will be the last important act of biological entities and if we get it wrong, "...the robots will eat us.". He further expresses his belief that robots will bring about the end time for humanity by exterminating the human race, but he is not concerned, he feels that the military robots are our offspring and at least they will survive.

You can view the guardians of your biological offspring's future at this site:
www.livescience.com/common/media/video/player.php

And find the 60-Minute segment concerning more Carnegie-Mellon mischief with thought identification, neural marketing and other mind reading robots here:
http://killerrobots.net/2009/06/30/robots-read-thoughts-directly-from-human-brain/

References:

"Autonomous Military Robotics: Risk, Ethics and Design"
http://ethics.calpoly.edu/ONR_report.pdf

"Future 'Bot: Robot-Human Convergence Begins"
www.livescience.com/common/media/video/player.php

60-Minute segment on mind reading robots
http://killerrobots.net/2009/06/30/robots-read-thoughts-directly-from-human-brain/