DEPENDING on what you read Artificial Intelligence (AI) is either the ultimate threat to humanity – always supposing we survive the climate crisis of course – or it’s our great saviour. Some argue that AI is developing so fast that it will take over the jobs currently done by humans, leaving humanity without meaning or purpose.
Others, while agreeing with that premise respond by saying ‘bring it on’. If the late David Graeber is to be believed, then many of the occupations we have now are nothing more than bullshit jobs that exist only because of the protestant work ethic – work is good regardless of how pointless it is. The argument runs that we should let robots get on with the jobs they are best at, and let humans do the work that robots can’t do – jobs that require compassion and judgement, for example. The second argument is often accompanied by the assertion that we need a Universal Basic Income to compensate for the reduction in paid labour, paid for by increased productivity from the robots.
Underlying the second argument there appears to be a more profound claim being made about the difference between humans and AI. And that is certainly the view of Brian Cantwell Smith in The Promise of Artificial Intelligence – Reckoning and Judgement. He starts the book by writing: “Neither deep learning, nor other forms of second-wave AI, nor any proposals yet advanced for third-wave, will lead to genuine intelligence.” And he draws a distinction between the brute reckoning of AI and the ‘human-level intelligence and judgement, honed over millennia, which if of a different order.
He reserves the word ‘judgement’ for the ‘normative ideal to which I argue we should hold full-blooded human intelligence – a form of dispassionate deliberative thought, grounded in ethical commitment and responsible action’. And although he acknowledges that not all human activity reaches this level, nevertheless it is an ideal to which ‘human thinking should ultimately aspire’ – an aspiration that is beyond AI.
Reckoning of the kind he attributes to AI refers to the kind of ‘calculating prowess at which computer and AI systems already excel – skills of extraordinary utility and importance’. Within this matrix Smith is most concerned that we humans in our admiration of AI attempt to emulate it by relying on ‘reckoning systems in situations that require genuine judgement’ and that by being ‘unduly impressed by reckoning prowess, we will shift our expectations of human mental activity in a reckoning direction’. Rather, he argues that we should indeed use AI to ‘shoulder the reckoning tasks at which they excel’ while we ‘strengthen, rather than weaken, our commitment to judgement, dispassion, ethics, and the world’.
A similar point was made by Prof Stuart Russell during his Reith Lecture on Radio 4 before Christmas. And although things don’t look great on issues like climate change, it is not all doom and gloom because these sorts of human responses have been successful in virtually eliminating the use of chemical and biological weapons.
Smith outlines four main areas in AI research that he thinks could counter the dangers inherent in systems that don’t exhibit these sorts of human attributes: 1 – take the body seriously; 2 – take context and surrounding situations seriously; 3 – consider the possibility that the mind is not just in the brain but extends into the environment as a form of ‘cognitive scaffolding’; 4 – don’t separate thinking from full-blooded participation and action.
Towards the end of the book, after a considerable amount of detailed analysis of AI, Smith returns to his main theme of distinguishing between reckoning and judgement, arguing that ‘ultimately, you cannot deal with the world…without judgement’. And that with judgement comes accountability which ‘can serve as the grounds not just of truth but of ethics as well’. Although he acknowledges that emotion, as well as reason, plays a significant role in our lives, he believes that the sort of judgement to which he refers – that is ‘authentic judgement’ – demands ‘detachment and dispassion‘ which can free us from the ‘very vicissitudes most characteristic of emotional states’. At the same time, however, he rejects the ‘idea that intelligence and rationality are adequately modelled by something like formal logic’.
This is a rich and rewarding book. And although Smith doesn’t refer to it, his position echoes Enlightenment thinkers who, properly understood, did not privilege Reason above all else, but argued that a little more reason in a world dominated by emotion, superstition and blind faith, might not be a bad thing.
Politically, it underpins a more communitarian approach over the individualism of liberalism. It is embeddedness in the world and in our communities that is the key feature in judgement and distinguishes is from the brute fact of reckoning.
Dickie's blog 