They have called it “revenge of the plebs”: masses of traditionally apathetic citizens becoming politically ignited and supporting populist politicians. The phenomenon gains traction across the western “liberal” world. In the US Donald Trump has caught the media and political elites off guard; coming from nowhere, with rhetoric that purposely deconstructs every semblance of political correctness, he’s now the President of the United States. In Great Britain, the referendum on whether to exit the European Union or not unleashes waves of dormant nationalism and xenophobia – as well as confusion and uncertainty. There is not a single country in the EU without a growing part of the electorate supporting eurosceptics, populists, and nationalists. “Plebs” have had enough of the ruling elites; they demand a catharsis of political life.
But what does this really mean? And what does it have to do with AI? Liberal democracy is based on a political idea that, in many ways, originates from Plato who in 5th century BC penned a very influential book called “The Republic”. In that book Plato suggested that the best political system is one where the wisest and best should rule, while the rest – the demos – should follow and obey. He broadly defined that ruling elite in terms of individuals who possess the necessary knowledge to run the affairs of the State, but also the moral integrity to be always fair and selfless. He did not support hereditary office: the ruling “best” could come from all strata of society as long they complied with the high moral and educational standards that were necessary.
Plato’s politics have had a tremendous influence in Western political thinking. They seem to uphold a reasonable tenet: that the affairs and decisions of government should be reserved for those who are morally and intellectually superior. Interestingly, Plato used the term “philosophers” to describe those ruling elites. The word ‘philosopher’ means ‘the one who is in love with wisdom’; and in today’s context it would mean scientists, engineers, economists and technocrats in general, anyone with the expert scientific knowledge that is necessary to run a government effectively and equitably. It makes sense: why shouldn’t we trust the running of our government to those with the appropriate skills and moral integrity to act impartially and unselfishly?
Where Plato’s argument fails is in the fallibility of human beings. As history and experience have consistently demonstrated, any human with the power to run other people’s affairs will most certainly place the continuation of that power as his or her top priority, well above the greater good. Ruling elites, once installed as is the case in representative democracies, will do everything to remain in power. Representative democracy is a system whereby an elite rules over the demos; and will therefore always have a “natural” tendency towards the collusion of elected officials and the rich. When the social contract between the demos and the elites breaks down – as it did during and after the Great Recession – the tendency unleashes a profound social rift, very much the phenomenon we witness today.
But what if those Platonic “philosophers” were not fallible humans, but infallible intelligent machines? What if their morality was neutral, and their intellectual ability several orders of magnitude higher than any human? Shouldn’t we let those benevolent, hyper-intelligent machines rule our world? After all, in the decades to come humans will become increasingly dependent on AI systems to run vital functions of their everyday personal, social, and economic life. Why not politics as well?
Many, of course, would rush to answer the previous question with an emphatic ‘no’. The idea of machines, benevolent or otherwise, ruling over human affairs seems like a non-starter. Nevertheless, if you think of the concepts of “civil service”, or indeed the “corporation”, these are non-human entities with agency (and legal standing) that act very much like “machines”. They run on processes, just like computers, they have inputs and outputs, and aim to be “impartial” and “rational”. Computers are simply facilitating a “government” or a “corporation” to execute its multiple “programs”. Adding machine intelligence to the already automated apparatus of government is the next logical (sic) step. Machine intelligence ought to eliminate human weakness and mischief from government. Given the choice, would you still prefer to be governed by corrupt politicians who collude with powerful industrial lobbies and make a mockery of our so-called democracy? Wouldn’t you rather choose reason and wisdom over persuasion and rhetoric? Given the evident alienation of voters from the established political class, this is a proposition that might appeal to a great number of citizens. Perhaps we should re-read Plato’s Republic as a prophecy predicting a future world governed by AIs.