Tuesday, 5 December 2017

Intelligence Explosion?

I can't resist passing on this piece (with a tip of the hat to Gareth Williams). It's a fine demolition of the increasingly popular notion that the use of AI (Artificial Intelligence) will at some point create an 'intelligence explosion' that will render us humans redundant and pose an existential threat to us. It's a scifi-inspired projection that rests on a fundamentally flawed idea of intelligence as a kind of superpower, a free-floating contextless phenomenon that, once it's let loose, will carry on doing its work until, as a result of exponential growth in competence, it has far surpassed our puny brains and will lead us into a future where AI is in charge and we humans are no longer needed. Intelligence – artificial or not – simply does not work like that, the author points out. We need not fear this 'explosion'; it will never happen.
Here is the link...

Of course, there may well be reasons to fear the effects of the spread of AI – not least the threat to jobs (this time, interestingly, to high-level as well as low-skill work). But AI isn't going to leave us all sitting around twiddling our thumbs, any more that the rise of digital technology did. AI will, I suspect, find its limits rather sooner than the more excitable futurologists predict. It will bump up against the bounds of the human world that created it and from which is cannot break free.

1 comment:

  1. Fascinating. The modern world is very primitive about this. It finds the fact of our duality almost impossible to encompass or deal with. I prescribe some John Donne.