AI could be the end of us as a species
Science fact has caught up with Science Fiction. Warfare has moved into a new phase.
The term robot is a contraction of the Czech word ‘robotnik’ which first featured in an SF play R.U.R. in 1920, written by Karel Čapek. It was followed up in Fritz Lang’s Metropolis as ‘Maschinenmensch.’ Philip K Dick explored the idea of automated weapons hunting humans based on their programming.
In 1942 the writer Isaac Asimov’ published his Laws of Robotics:
· First Law: A robot may not injure a human being or, through inaction, allow a human being to come to harm.
· Second Law: A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
· Third Law: A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
He later added the ‘Zeroth Law’ (i.e. ranking higher than the others): “A Robot must not harm Humanity.” Asimov wrote a number of novels and short stories where he explored the implications of these laws. Other writers have developed these themes. Robots have always offered humanity the dream of a life free from labour exploitation. This has always been tempered by the threat of robots turning on their human masters.
The real and present danger is that in the Donbas, both Russia and Ukraine are using automated drones as active combatants. Ukraine recently claimed that its drones completely eliminated a Russian infantry unit entirely without human assistance.
Most of these drones are FPV (First Person Viewer), whereby a remote human operator controls the drone in the same way as they would play a role-playing video game (RPG). Ukraine has made massive strides in the development of this technology, with the involvement of major tech companies who are using the war for beta-testing.
However a critical red line has already been crossed. Due to problems communicating with these weapons while undergoing severe jamming signals, remote operators are switching to autonomous systems powered by AI. Once the target has been identified and ‘locked-in’ to the system, the AI does the rest, even if this means directly or indirectly killing humans.
And then the feedback loop closes. The data of humans being killed by machines becomes part of the training data set for future AI, any attempt to impose something like Asimov’s Laws will be treated by AI not as commandments but as data to be evaluated against other data sources. The trap is sprung.