Just Finished Reading: The Most Human Human – What Artificial Intelligence Teaches Us About Being Alive by Brian Christian (FP: 2011) [271pp]

The Test existed, at least on paper, long before there were any contestants even theoretically capable of taking it. In 1950 mathematician and early computer scientist Alan Turing devised a way of determining if a machine was exhibiting intelligent behaviour. If it could communicate with others without them being aware that they were talking to a machine, rather than an actual flesh and blood person, then at least in theory it could be argued that the machine, computer or software programme exhibited recognisable intelligence. Essentially, in order to pass the Turing Test and win any prize on offer, the idea is to create a machine that can imitate human communication (hence calling it the ‘Imitation Game’) enough to fool a set of interrogators over a 20 minute period. The rules have varied over the years but, as yet anyway, the Turing Prize has yet to be won. But that’s only one aspect of what’s going on in Turing territory – as the author discovered.

The present Test has another aspect that I hadn’t known about until picking up this fun and fascinating book. Apart from the prize given to the Most Human Machine there is another given to the Most Human Human. As part of the test the set of interrogators ‘interview’ both the machine competitors as well as a few actual humans who’s remit is the act as human as possible. The author, coming from a background in both computer science and philosophy, knew that wasn’t anywhere near as easy as it sounds – so he decided to ‘train’ for the event by speaking to experts in the field of language, communications, chess programmes (including some world class players who are not amused at all at being beaten by computer programmes on a regular basis – so much so that many people the era of competitive chess is dead), video-game design, speed-dating, criminal law, neurology and psychiatry. It was an interesting ride! I’ve always had an interest in how we ‘read’ other people and especially how we can tell if other people are being less than honest with us. How the brain does that and how e-mail scammers try to circumvent that ability was just one of the avenues the author explored to help him ensure victory for humankind.

We increasingly live in a machine world – a world built by machines and, increasingly, a world built for machines. Algorithms help us pick our next book, our next movie or next meal. Algorithms increasingly advise us on who should be our friends and what news we should listen to. They do this, at least in part, by parsing what they can understand as ‘human nature’ into a digital format that can be processed, manipulated and fed back to us. Ironically it could be argued that the Turing Test is becoming easier year on year, not just because computers are becoming more powerful or algorithms are becoming more sophisticated but because we humans are becoming more machine like in our thoughts, our actions and our communications. The author, rather brilliantly, decided to go the other way and to display everything that makes human communication and humanity itself so interesting, so nuanced and so much fun (and so frustrating and a complete pain in the ass at times!). Definitely a unique look at AI, future tech and the world we’ll all be inhabiting much sooner than many of us expect and a guidebook to keeping hold of our humanity as we get there. Recommended.