This morning’s Observer [column][(https://www.theguardian.com/commentisfree/2021/feb/06/mark-zuckerberg-tim-cook-facebook-apple-problem):

A few years ago, during a period when there was much heated anxiety about “superintelligence” and the prospects for humanity in a world dominated by machines, the political theorist David Runciman gently pointed out that we have been living under superintelligent AIs for a couple of centuries. They’re called corporations: sociopathic, socio-technical machines that remorselessly try to achieve whatever purpose has been set for them, which in our day is to “maximise shareholder value”. Or, as Milton Friedman succinctly put it: “The only corporate social responsibility a company has is to maximise its profits.”

Given that, it doesn’t really matter whether those who sit at the top of giant tech corporations are saints, sinners or merely liars and hypocrites. Facebook could be staffed entirely by clones of St Francis of Assisi and it would still be a toxic organisation, relentlessly pushing to achieve the purpose assigned to it by Professor Friedman. So if we want to make things better, our focus has to be changing the machine’s purpose and obligations, not on trying to persuading its helmsman to attend to the better angels of his nature.

Later. The point about corporations being sociopathic, superintelligent machines prompted an interesting email from Ross Anderson (Whom God Preserve) gently pointing out that he had made a similar point to the philosopher Daniel Dennett during a lecture Dennett gave in Cambridge in June 2019, which of course led me into an interesting dive back through diaries and notebooks.

Tracing the genesis of an idea can be a fool’s errand, especially if it’s a powerful idea. But this one proved more fruitful than I expected.

On the chronology, the Dennett lecture that Ross mentioned was in June 2019, but I had comeacross the ‘sociopathic’ idea before then in one of four seminars that David Runciman ran for the Centre for the Future of Intelligence in 2017-8 and I attended. It came up because he’s an expert on Hobbes and he argued at one point that Hobbes’s Leviathan was, essentially, an AI. The political philosopher, Philip Pettit, who was also at the seminar took up the idea and argued that corporations are intrinsically sociopathic — which is what I took away from that particular conversation (though I may have been wrong in attributing the sociopathy attribute to David rather than to Philip).

But it seems that the question of corporate sociopathy was been around a long time in American constitutional discourse. In part this seems to have been about the (ancient) legal doctrine that companies were granted “legal personhood” long ago. In 2003 Joel Bakan, a legal scholar, published The Corporation: The Pathological Pursuit of Profit and Power, a book which was accompanied by a documentary film, The Corporation (which I haven’t seen — https://www.imdb.com/title/tt0379225/) but which gave rise to lots of commentary about psycho/socio-pathy (e.g. https://www.psychologytoday.com/us/blog/machiavellians-gulling-the-rubes/201605/are-corporations-inherently-psychopathic). And then, of course, there was the 2010 ‘Citizens United’ judgment of the US Supreme Court which held that the First Amendment prohibits the government from restricting independent expenditures for political communications by corporations, including nonprofit corporations, labor unions, and other associations. Which effectively means that in the US these sociopaths have free speech rights and can fund politicians who advance their interests.

I’m sure there’s a lot more but the bottom line is that this is an older idea than I had thought and it’s been been around for quite a while.