Would the world be a darker place ruled by robots or are they really our knights in shiny armour? Humanity at the mercy of technology has been a core theme of science fiction for decades, but over the years audiences have seen the origins of this threat gradually shift. From the far reaches of “Forbidden Planet” and “The Day the Earth Stood Still,” more recent cinema has tended towards presenting a more home-grown menace of Skynet in “The Terminator“, “The Matrix” and even “I, Robot“. But is there really anything to fear? Well, according to the late, great, robotic-voiced Professor Stephen Hawking:
“The development of full artificial intelligence (A.I.) could spell the end of the human race.”
This worrying claim is based upon the certain notion of arriving at a point in the not too distant future known as “the singularity” – where A.I. matches and then surpasses human intelligence. At this point, would a machine that can outthink its human master free itself from the shackles of humanity and become something much more powerful (perhaps, even better) than us? Dr. Joanna Bryson doesn’t think so.
A specialist in A.I. at the University of Bath’s Department of Computer Science, Dr. Bryson argues that A.I. and machine learning should not be seen as an intelligent entity in its own right. Bryson thinks we should view it as a tool or extension of humanity which, if managed correctly by us, could offer a huge number of benefits to our world. We are already beginning to see this in intelligent systems that analyse health and environmental data that raise early alarms, machines that operate and respond in dangerous and emergency situations, technology for more safe and efficient self-driving transportation, as well as ‘smart’ information systems such as Siri, Google and Bing.
But to what end? If we are already happy for machines to manage the trivial for us, then why would we not sacrifice an even greater amount of control to systems that operates purely on intelligence and logic? Would it not be more beneficial for us to be governed free of often controversial constraints such as religion, politics, greed, and war? Dr. Bryson doesn’t think so. Not only does she view it as really nihilistic to entertain the idea of machines living for us, but in fact, far from becoming more removed from humanity, Dr. Bryson has presented research that demonstrates A.I. has even started adopting our existing human biases. Her analysis of search engine results revealed that some seemingly neutral search terms deliver results with a greater association of a particular gender or race, for example “technology” often being associated with men and “unemployment” regularly being associated with African Americans. Further examples of A.I. learned bias has also been observed by others, such as ‘Tay’ – a Microsoft developed chat-bot, designed to talk on Twitter like a teenage girl that eventually became obsessed with sex, herself and Hitler.
So, overall, it seems that we’re actually unlikely to see the mechanical colossus of 1950s drive-ins crashing around our cities seeking to “DESTROY ALL HUMANS!”. Bryson believes that it is impossible for A.I. and machine learning to detach itself from the inherent flaws of its creators – humans. This means that it’s all down to us to create a better future – if that is what we want. It is us, as a species, that has the true power, and it is us that needs to learn and grow if we are ever going to be able to fully harness the positive power of the innovative, life and world-altering technology of A.I. Perhaps that is even more scary a thought.