The other night I went to Ben Hammersley’s talk at the brilliant Royal Institution in Mayfair. The overarching (overbearing?) theme – The Robots Will Kill Us All – AI and Post Digital Geopolitics – actually turned out to be relatively lighthearted considering – mostly thanks to the erudite delivery by Hammersley.
We kicked off with checking out some emergent behaviour from self correcting, simple systems – how they make themselves more attractive to one another to swarm into an emergent group – see swarm-bots.org and the rather portentous video of a swarm of bots making off with a child above.
We moved on to find out how the collective might of these bots is being applied in the world of finance. Take the innocuous sounding Hyde Park Global – one of a few who have developed genetic algorithms whose fitness is to produce more profit. There are no traders per se because the stock market is being played by AI – and these newsflow AIs get about a 20 minute headstart on the human trader capability. A few years ago NY saw a flash crash where these AIs started to freak at one another and began selling spontaneously. It was this moment that these AIs made their debut into the day to day vernacular, popping up above the parapet to become known to the layman.
Most interestingly and positive was the learning that Google is working with the World Health Organisation to predict flu epidemics: they can spot unusual behaviour and searches in their mass of data right across the world. WHO can now react to this nascent intelligence.
