Machine consciousness, sentience and mind

When technology becomes ubiquitous, bad habits start to form, marked by data breaches, identity threats etc. and the power of AI can be grossly misused


Artificial Intelligence, the term, was coined way back in 1956 by John McCarthy, a Stanford professor. Photo: iStock
Artificial Intelligence, the term, was coined way back in 1956 by John McCarthy, a Stanford professor. Photo: iStock

Artificial Intelligence, the term, was coined way back in 1956 by John McCarthy, a Stanford professor. As an idea it had its share of disappointments, battled scepticism and was kept on the backburner for several decades. However, intelligent machines and for them to be considered at par with human intelligence, are but two different propositions. Sci-fi readers would recollect HAL 9000, Arthur C. Clarke’s AI–based protagonist in the book 2001: A Space Odyssey, which turns to an antagonist as the plot unfolds, and eventually turns villainous. Will it remain fictional, or is there a strong possibility that we may actually experience this in our lifetime?

Narrow AI

Now AI, in its less dramatic form already exists. Task driven or “Narrow AI” is about focussing software-driven cognitive intelligence on a specific task within a limited context. Unconsciously, we interact with Narrow AI all the time—credit card fraud detection, speech recognition, web search rankings, automated customer service are common enough examples. Two billion smartphone users, and that 40% of the global population is online, act as catalysts to accentuate progress.

Personalization is a rage, and its rapidly growing popularity is because of Narrow AI. It is transforming the IT BPM industry as well. AI-led automation and augmentation across service lines, is likely to deliver $100–120 billion of net productivity gains by 2025. Translated in revenue terms, the next 7-10 years will usher in an additional $400 billion due to the industry’s relentless pursuit to get into the “heart of the business.” The caveat here is about picking the right battles and not yielding to the hype. Thumb rule proffers—the impact should be at least 5-10 times.

Alan Turing, the famous mathematician of the last century, had propounded a theory. While interacting with a human, if a machine can fool him/her into believing that it’s another human, then it can be termed as intelligent. Yet again, this theory encompasses Narrow AI or Weak AI, and it does not traverse into the deep realms of General Purpose or Strong AI.

Cognitive Technologies and the march towards Strong AI

Advances in AI is due to knowledge synthesis, self learning, context awareness and natural communication. How soon we will see wider application of Strong AI, would be a factor of the progress made in these areas, including the quantum of funding that will come about.

In 1952, the computer learnt how to play Tic-Tac-Toe. In 1997, Deep Blue beat Gary Kasparov, the reigning world champion at that time. In 2011, IBM Watson wowed the tech industry with its historic win against two of television quiz show Jeopardy’s greatest champions. Hailed by the epithet of “Jeopardy-winning machines,” Rutter and Jennings were touted to be unbeatable. Between them, they’d racked up over $5 million in winnings on Jeopardy, and in early 2011, they agreed to an exhibition match against an opponent who’d never even stood behind a Jeopardy podium before. IBM Watson won!

“I, for one, welcome our new computer overlords,” was Ken Jennings’ response after the loss.

AlphaGo, the board-game-playing AI from Google’s DeepMind subsidiary, is one of the most famous examples of deep learning—machine learning using neural networks—to date. The software which taught itself to play the ancient game of Go by running millions of simulations against itself, eventually defeated the Korean Champion. It is an example of supervised learning (studying previous games played by humans) and reinforcement learning (playing against itself and learning from the mistakes).

The Second Wave—the Dark Side of Digital?

Artificial Intelligence can actually help us solve some massive problems in the areas of energy, pollution, clean water, for instance. A golden age is conceivable, if it remains in the right hands. However, in the 2nd wave, bad actors cannot be avoided. When technology becomes ubiquitous, bad habits start to form, marked by data breaches, identity threats etc. and the power of AI can be grossly misused. The foundation of software will be expanded by them with an intent to cause explicit harm. As Elon Musk prophesizes, “it could be like summoning the demons.”

Would it be a stretch to imagine a future when machines in their bid to get faster and more accurate, start viewing humans as “obstacles to efficiency”?

Then of course is the whole debate about job losses—how many, and how real is the concern? Millions, if various reports are to be believed! Some even opine that 47% of jobs will directly come under threat. This time however the rate of displacement will be very high, a phenomena we have never experienced before! Notwithstanding overwrought pleas from Luddites, it is but technology, which can bridge the yawning gap between haves and have-nots.

Raman Roy is vice-chairman at Nasscom and chairman and managing director of Quatrro Global Solutions.

More From Livemint