by Adam Yardley
Artificial intelligence, or AI, is always on the advance – it’s incredible to think that machines are now outstripping people on many different levels and at many different tasks – and as AI research only advances further, things are starting to get a little scary. How far are we away from the sci-fi staple that is the ‘robot uprising’? It perhaps won’t happen today, tomorrow, or even next year – but AI is now able to learn more about people than ever before – and it can do so simply by looking straight into their eyes. They often say that the eyes can give a person away – and that’s never been truer.
A global research team led by Tobias Loetscher, at the University of South Australia, put new technology to the test in looking for basic personality traits and triggers which could be given away by certain eye movements. With help from universities in Germany, the team involved were able to find that they could find at least four personality traits by just analysing eye movements and patterns – and while their findings may not be taken as red just yet, it’s stunning to think that a machine can now know more about you than ever before – without you having to utter a word.
The machines used in the study of 42 participants were found to be able to ascertain neuroticism, conscientiousness, agreeableness and extroversion – the only element the AI reportedly had problems with was that of openness to new experiences – as this, largely, may differ from person to person. It’s thought that this study may provide a new in-road for software developers looking to further personalise their services and standards moving forwards. “People are always looking for improved, personalized services – however, today’s robots and computers are not socially aware, so they cannot adapt to non-verbal cues,” Loetscher advised. “This research provides opportunities to develop robots and computers so that they can become more natural, and better at interpreting human social signals.”
It’s certainly impressive stuff – our eyes and facial patterns are already being used in line with a range of software and hardware – but what does this new study mean for further personalization moving forwards? Could this mean better-tailored services for us all – or could it mean a whole new set of privacy concerns lying in wait in just a few years’ time?