Editor's note: The possibilities and dangers of an AI-emboldened world

The machine only gives us numbers because I would rather that a human element remain in determining something so critical as someone’s fate.”

That’s one of the lines of Harold Finch, a reclusive billionaire computer nerd, in the American sci-fi crime drama television series Person of Interest that aired on CBS from 2011 to 2016. Finch has developed an artificially intelligent system for the US Federal government, known as ‘the Machine’, which can sift through and analyse data collected by the National Security Agency’s mass surveillance to accurately predict terrorist acts and people planning them. Aside from being a splendid crime-fighting show, the series gained acclaim for raising issues inherent in the use of artificial intelligence (AI), from privacy and the power derived from social surveillance to the moral and ethical issues of human control and the problems of working with limited information programs.

In this special issue on AI, we dive into the possibilities and the dangers of an AI-emboldened world in a fortnight in which disclosures from a global journalistic collaborative about a software meant to combat terrorism and crime suggest a different technology may have been used to create the most sophisticated smartphone attack ever. The makers of the software, an Israeli cyber arms firm called the NSO Group, have denied the allegations, including that data was leaked from their servers, that their technology was used to collect information on assassinated Saudi Arabian journalist and dissident Jamal Khashoggi, and that the software was used to hack the phones of some Indians, including journalists.

The good news, though, is AI technologies go way beyond surveillance, facial recognition systems and smart policing. It doesn’t necessarily follow that all countries that do use AI are abusing it for mass surveillance.

As Forbes India’s Technology Editor Harichandan Arakali writes, the coming together of cloud computing and small devices that can perform enough computation at the extremities of connected network is aiding the spread of AI-based solutions in multiple sectors.