robocop_rolandakirolandAutonomy has been a central issue in philosophy for millennia. Traditionally it is the cluster of concepts and problems concerning Freedom, Free Will, Determinism, and Fate. Kant re-focused this conceptual cluster to the question of autonomy or “self-directedness.” The Existentialists developed this theme of self-direction into the 20th century [1.]

Now in the 21st century we have the very real prospect of autonomous technology: robots, computers, neural nets, webs, and machines that act on their own based on environmental factors and programming. It is here now.

While the debate over genuine Artificial Intelligence (paralleled by the debate over what human intelligence is) continues, we have self-driving cars on the highways and self-flying drones over our cities and battlefields. The practical questions before us now are not “is autonomous technology possible?” but instead, “should we arm autonomous drones and tanks to kill and should we allow autonomous medical robots to prescribe and apply treatments?”

Even those questions are moot now given that Israel and South Korea have M2-Catlanit002deployed autonomous lethal guard robots. The robocat is out of the bag and we may expect an autonomous arms race to ensue from here on, perhaps until some horrible catastrophe occasions international reflection [2, 3.]

For me it is hard to imagine just what would really cause us to think and act to limit autonomous weapons. After all, we in the West shrug off the assassination of an entire wedding party (17 people killed) by a drone missile [4.] Exactly what would get our attention?

I understand that some people will rise to the defense of such collateral damage. That’s not my issue here (though it is elsewhere). My interest here is in what we are individually and collectively willing to allow autonomous technology to become, even if the prospect of it’s development threatens the future of humanity as a whole.

Steve Ranger addresses these concerns in; Robots of death, robots of love: The reality of android soldiers and why laws for robots are doomed to failure [7.] His thesis is that we are creating and deploying autonomous technologies that can change our lives without our consent and we are doing so long before any level of moral reasoning is available to those machines. That is, we are enabling self-directed machines to make life and death decisions for humans when the machines have no concepts or mechanisms for recognizing and understanding the values of death and pain.

There is a rich discussion in philosophy and engineering over the conditions of safe and ethical artificial intelligence [8].  The issue at present is not AI, it is automated unintelligence. It is a lively debate over whether AI will save us or destroy us, but there much speculation needed to recognize the malicious effects of weapons programmed to kill automatically without intelligent judgment.

Pushing the dilemmas of autonomous technology further are thinkers who ai_artificial_intelligencequestion the relevance of the human species in world with autonomous artificial intelligence. James Barrat’s recent book; Our Final Invention: Artificial Intelligence and the End of the Human Era is a thoughtful analysis of such a pessimistic (for humans anyway) future [9.]

Of course, there are plenty of fictional depictions of autonomous technologies. One of the more reflective – in a fairy tale way – is A.I. Artificial Intelligence directed by Stephen Spielberg based on a script by Stanley Kubrick [10.] I’ll be very interested in your thoughts about this film and these fascinating, in some ways disturbing, philosophical topics.

Sources
1. Buss, S. (2013). Personal Autonomy. Stanford Encyclopedia of Philosophy.
http://plato.stanford.edu/entries/personal-autonomy/

2. Shachtman, N. (2008). Israeli “Auto Kill Zone.” Wired.
http://www.wired.com/dangerroom/2008/12/israeli-auto-ki/

3. Page, L. (2007). South Korea to field gun-cam robots on DMZ. The Register.
http://www.theregister.co.uk/2007/03/14/south_korean_gun_bots/

4. Ali & King. (2013). Fatal air strike hits wedding convoy in Yemen.
http://www.latimes.com/world/worldnow/la-fg-wn-yemen-drone-strike-wedding-20131213,0,4137970.story#axzz2o358s0vm

5. Department of Defense Directive 3000.09: Autonomy in Weapon Systems. http://www.dtic.mil/whs/directives/corres/pdf/300009p.pdf

6. Yudkowsky, E. (2013). Five theses, two lemmas, and a couple of strategic implications.
http://intelligence.org/2013/05/05/five-theses-two-lemmas-and-a-couple-of-strategic-implications

7. Ranger, S. (2013). Robots of death, robots of love: The reality of android soldiers and why laws for robots are doomed to failure. TechRepublic.http://www.techrepublic.com/article/robots-of-death-robots-of-love-the-reality-of-android-soldiers-and-why-laws-for-robots-are-doomed-to-failure

8. Yampolskiy, R. and Fox, J. (2013). Safety Engineering for Artificial General Intelligence. Topoi, 32:217–226
http://link.springer.com/article/10.1007%2Fs11245-012-9128-9

9. Muehlhauser, L. (2013). Our Final Invention: Artificial Intelligence and the End of the Human Era.
http://www.kurzweilai.net/book-review-our-final-invention-artificial-intelligence-and-the-end-of-the-human-era

10. Spielberg, S. (2006). A.I. http://www.youtube.com/watch?v=sqS83f-NUww

Image acknowledgements:

Robocop, Roland Molnár, https://www.flickr.com/photos/rolandakiroland/8935247685/

M2-Catlanit002, Wikipedia
http://en.wikipedia.org/wiki/Samson_Remote_Controlled_Weapon_Station

Advertisements