In the conduct of war, military leaders always have the incentive to mitigate the physical and psychological harm caused to their own personnel. The use of aerial bombing is the means by which technologically sophisticated nations use to inflict as much harm as possible on an enemy with as little risk as possible. Unmanned Aerial Vehicles are a new and poorly understood development in warfare that allows for a nation to be more discriminate in who is actually killed, than conventional manned aircraft. They allow for longer loiter times over a given target and are equipped with various sensors and cameras for teams of personnel to dissect the situation on the ground and initiate the strike.
Theoretically this should be relatively uncontroversial. If a war is to take place, then it is better to conduct it in a way that reduces harm to all innocent parties involved. However, remote killing is not so simple and it does allow for large margins of error. Signature strikes guided by rudimentary machine intelligence programs which flag potential targets via meta-data through cellphones, and visual intelligence software that looks for 60 patterns of movement to determine if an individual is a terrorist, are some examples of where technology may be trusted beyond the extent for which a decision making human can understand the potential error. Through leaks, the public can now see that in one month, 90% of people killed were not the intended targets. This can be summed up to be a terrorist campaign, as there is very rarely an explanation and it is often in countries with no formal declaration of war.
As we proceed into the future, entire career fields will be replaced by autonomous decision makers. This will be done in secret, and it will most likely come home. This talk is intended to provide the public with the perspective of a whistleblower and to prescribe actions they might be able to take to restrict this development.