Many people would be familiar with the name Alan Turing from the 2014 historical drama, The Imitation Game, which was the highest grossing film of that year. Turing is widely considered to be the father of artificial intelligence and I remember studying one of his concepts at University. It was known as the Turing test.
This is a test of a machine’s ability to exhibit intelligent behaviour indistinguishable from that of a human. The test came from his paper that posed the simple question: “Can machines think?” Turing went on to explain that ‘thinking’ is difficult to define so posed the solution to the problem with the idea of a game called the ‘imitation game.’ If a person has a text only conversation between either a computer or a person and can’t reliably tell the machine from the human, then the machine has passed the test.
Turing’s paper was written in 1950 when the majority of communications occurred verbally – mostly in person. Telephones were available but, in the US, there was only one telephone for every five people. Some at the time dismissed the Turing test as not a relevant test of ‘thinking’ because it removed voice from the conversation and relied on a text only conversation. When would humans have a conversation just with text?
Fifty-six years later along came Twitter. The perfect platform for Turing’s test. Along with texting and e-mail and other social media platforms, think of how many people you have had conversations with that have been purely via text.
Luckily we know that we are having those conversations with real people and we could definitely tell the difference if it was just a computer at the other end.
Or could we?
Researchers have found that bots are used extensively on Twitter to amplify a message. 38 per cent of tweets related to ‘fake science’ were from bots. 28 per cent of tweets related to Exxon were not from humans and 25 per cent of climate change denial tweets were bot generated. An increase in social media traffic is a way that many people, and even media organisations, determine that there is public interest in a particular topic. It is a principle known as ‘social proof.’ It is why we are attracted to a restaurant that seems busy despite the fact that we would probably receive our meal quicker in an empty restaurant. If lots of other people are eating there, it must be good! Right?
If an organisation created a method to influence public opinion just by tweeting more, surely that could have a major impact on the bottom line. But it would be unethical so that would never happen. Think again. It was discovered last year that the mining company, Glencore, spent up to £7 million per year with the C|T Group over several years with teams in Sydney and London to spread anti-renewables messages and counter anti-coal activists.
The latest evidence shows that Glencore was charged way too much. Other companies are using bots – at a much cheaper rate – to deliver the same social media influencing.
There is no doubt that Turing was a visionary and way ahead of his time but I am not sure that even Turing could have imagined that his theoretical test would actually be used every day by billions across the world.
Next time you read comments or interact with someone on social media that you have never met, ask yourself if you think it is actually a human or a bot – and then how you really know.
Send any further thoughts to ask@techtalk.digital – humans only please.
Mathew Dickerson