By their own admission, Jack Ma is uncomfortable with Artificial Intelligence (AI) and Elon Musk is scared. But why? Contrary to popular perception AI is old. To be precise it’s 51-years old, widely acknowledged to have been born at a conference at Dartmouth College in 1956.

That conference was attended by a diverse group of people. Three of them presented the Logic Theorist, the world’s first true artificial intelligence programme. Two of them were Allen Newell and Herbert Simon. Both were not typical computer scientists or programmers. Newell was a physicist trained in mathematics and who became deeply interested in cognitive psychology. Simon was Newell’s teacher and advisor. He was a political scientist and economist and moved laterally into sociology, management sciences, philosophy of sciences and cognitive psychology. He received a Nobel Prize in Economics and the coveted Turing award. John Clifford Shaw was the junior partner and lone programmer, and as Newell himself admitted “the genuine computer scientist of the three”. Such was the optimism about AI that Simon boldly predicted that “machines will be capable, within twenty years, of doing any work of man”. Of course, like several predications of the future driven more by the all-round hope and enthusiasm of the 1950s rather than on the complex realities, this one too fell far short of its promise.

AI proved to be far more complex and complicated, simply because it sought to replicate human cognition by seeking to convert it into discernible frameworks of rationality, logic and decision making. Humans are genetically tuned to be instrumental and abstract, local and global, tactical and strategic, and all often at the same time. Eons of hunting, gathering, protecting against the elements and other beings, stocking up for the future and playing the extremely nuanced mating rituals have turned human beings into a species that has distilled collective self-interest into an intuitive way of life. We refer to it as common sense; a logical underpinning widely accepted and understood as a cognitive framework for all human action and activity. Simon’s bold prediction proved off target because AI programmers just couldn’t replicate all possible permutations and combinations of common sense that informed general human decision tree into a replicable and predictable framework.

The real possibility of AI mastering all dimensions of conventional human rationality revived when IBM’s Deep Blue took on the chess world in 1996. Deep Blue and its more powerful version informally called Deeper Blue was just brute computing power and was not AI. At its peak performance, Deeper Blue was capable of evaluating 200 million positions per second. The clear tipping point for AI was Deep Blue’s victory over the maverick chess genius Garry Kasparov, possibly the greatest player the world has known till now. Such was the impact of the victory on Kasparov that he saw “superior, deep intelligence and creativity” in the moves of the machine. Though the Deep Blue machine was mothballed and shelved for various reasons, Kasparov’s defeat by the machine still did not trigger alarm bells among technocratic and corporate elite. The seeming mastery of a machine over a human was after all within the confines of 64 squares. A world filled with permutations and combinations, yet still finite. The internet world, ironically, in its search to humanise its interaction and engagement with its audience catalysed a different kind of software programming where the coding effort was oriented towards capturing user behaviour in order to serve them better. Today, we know it by different names, from customisation, personalisation to big brother and surveillance, depending on our inclinations and worldviews. The all-pervasive envelope of the software systems made a lot of us uncomfortable, but the questions were more on the lines of individual rights, privacy, data security and cybercrimes. The nature of the software and the questions that arose from it were still within the norms and frameworks of typical human rationality. In short, both the scope of the software and the questions emerging from it could be contained with active human intervention in the form of regulations, policies, explicit consent and other filters.

Yet, the most sophisticated of these software systems today are not true blue AI. They are really smart pieces of codes that are amalgamated to recognise patterns and predict them in advance based on the moves that we make. The software system mimic human behaviour at its stereotypical best: patterns driven by personality types and triggers. It’s cutting edge, no doubt, with many of these software systems self-learning, but they are still within the norms of predictable rationality. In short, the software can still be written and rewritten by human beings. But what makes human beings, well humans, is our epiphanies, rebellions, counter-intuitions and the lone walks. Humans are humans because of an embedded unpredictable rationality that’s always on the horizon. You never know when it will pop up. It’s Pablo Picasso and Nikola Tesla. It’s also Adolf Hitler and Idi Amin. It’s also Mahatma Gandhi and Nelson Mandela.  Neurons in the human brain are being wired and rewired in such unique and diverse ways autonomously and continuously to all sorts of external and internal stimulation that each one of us is a self-contained and self-looping system. In short, each human is sentient and when humans interact with each other as sentient beings there is larger and more powerful collective intelligence.

Jack Ma is uncomfortable and Elon Musk is scared because Artificial Intelligence, or software systems written by us humans, have crossed the threshold of sentience, the holy grail for any software system to considered as true blue AI. It’s not in the nature of Jack Ma to be alarmist, but he did say, “Thirty years later, the Time magazine cover for the best CEO of the year very likely will be a robot. It remembers better than you, it counts faster than you, and it won’t be angry with competitors.”  The most prominent AI system is used by Musk’s Tesla Electric Cars to navigate itself through the city and make decisions normally made by a human being. What’s more, the software keeps coding and recoding itself depending on the scenarios and situations that it finds itself in. Much like the human brain. It can also network with other such software systems, exactly like how humans connect with each other to create collective intelligence. Now that the sentience threshold is well and truly crossed, the only fence holding AI back is one planted by humans. It won’t be long before that fence comes down. Ma and Musk are scared because human beings for the first time are confronted by a real possibility of losing control over themselves. But should they be?

Swaminathan is a digital native and has lived through three dotcom bubbles and busts.

(The column appears in the May 31, 2017 issue)