What do AI technologies ‘see’?

Artificial intelligence (AI), also known as machine intelligence, is a term used to describe machines (such as computers) which mimic ‘cognitive’ functions that humans associate with the human mind, such as ‘learning’ and ‘problem-solving’. Nowadays, AI, including algorithms, software and smart technologies, have a rapidly growing presence in cities across the world: AI technologies can be found all over cities, from lampposts to rubbish bins, traffic lights and cars.

A recent study which sought to explore the rise of smart city technologies, examined ‘how computers see the public space and the effect of algorithms, artificial intelligences and automated processes on the human experience in public spaces’. The lead researcher, Silvio Carta, aimed ‘to see what machines see, […] to think like computers’. To do so, he created a simulation, The Machine Eye, which demonstrates the steps through which a hypothetical AI system ‘reads’ a physical environment and is able to profile the people in it.  He suggests that it starts off with ‘a pitch-black situation – with no information – and gradually gathers data from a number of interconnected devices’ including smartphones, microphones, CCTV and other sensor. By bringing together all the data captured, AI machine can construct accurate profiles of targeted individuals.

Students considering applying to Computer Science should examine and assess the ever-increasing rise and development of such AI systems and machine learning processes, and reflect on how they are ‘essentially mathematical and statistical models, cleverly applied to real-life problems’, thus demystifying these so-called ‘black-box technologies. Those wishing to apply for Politics may also consider this topic in dire need of discussion: in an age of electronic communications, mass surveillance and pervasive security threats, the modern state has amassed far more data and knowledge of its citizens’ behaviour than ever before, the consequences of which may be unsettling if mishandled or inappropriately managed.