Artificial Intelligence - A Gift Or A Curse?

It started with an unpretentious classification of a branch of Computer Science, “to make computers behave like humans”, in 1956. Its first public ‘aha!’ moment was when Deep Blue, an IBM Supercomputer, beat Gary Kasparov, in their rematch of 1997.

The final game was a shocker. Deep Blue played an aggressive game, sacrificed a Knight, and defeated the Grand Master in 19 moves. The world sat up and looked, but felt warm in the comfort that Artificial Intelligence (AI) was indeed a great thing for people to play against. But just that.

On December 7 Dec 2017, Google Alpha 0, defeated Stockfish 8, in a series of chess games. A match between Super AIs. Significant is the fact that Stockfish 8, had access to decades of computer experience, trained on centuries of chess moves.

It could calculate 70 million chess positions simultaneously. Alpha 0 could calculate 80,000 positions simultaneously and had never learnt Chess. It was allowed 4 hours of preparation for the match.

Of the 100 games played it won 28, drew 72, and went from ignorance to mastery in 4 hours. This changed the paradigm for AI. Alpha 0 could learn on the job. Machine Learning is based on deep neural networks in the AI, similar to human learning.

Two years later in 2019, Pluribus, an AI from Carnegie Mellon, was fielded to play multiplayer Poker. It played 5000 hands against 12 professionals, five at a time and won. It used mixed strategies, unpredictability, and exceptional computing to engineer its wins.

Clearly, the minds developing AI were on track. The wider body of intelligentsia, however, debated the unlikelihood of computers ever reaching the sophistication and complexity of the human brain.

By this time, pre-COVID, AI was powering facial recognition, computer speech, speech recognition and of course Robotics. Sophia, a social humanoid robot was developed by Hanson Robotics, Hong Kong in 2016 and became a UNDP-recognised entity in 2017.

As Sophia effortlessly appears on TV and gives interviews, we need to remember that it combines 17 Key AI technological disciplines and uses 1000 solvers and path planners, just for locomotion. In Feb 2022, appeared Nikola, a child Android, which can emote, and convey happiness, surprise, sadness, fear, anger and disgust.

Let’s slip into the domain of our ordinary lives. Music apps, gaming, web shopping, searches on the net, food delivery apps, taxis, air ticketing hotel bookings are fired by effective AI.

How often have you listened to music that you want, or just clicked into the curated playlist innocently labelled ‘especially for you’ or some such seductively innocent tag? Try buying clothing online more than once. And of course, do experience online air ticketing. Wander a bit and come back to an earlier option and Bingo, you get a price rise.

The finest artificial minds are at play, manipulating choice, eking out every possible avenue of profit, and capturing your cognitive space. The impact of AI, on the creative and performing arts is well known. From the agitations in Hollywood to the acrimonious debates in the Art world AI, has roiled the domain.

On Nov 17, 2023, the board of Open AI, removed Sam Altman, and subsequent to five days of feverish developments, he was reinstated, with members of the Board being terminated, instead. While the facts await an inquiry, and may never be known, significant aspects have emerged.

Open AI’s Board followed an unusual structure, with an unpaid, non-profit Board, having a charter to oversee a major AI Start-up. Central to its charter was the need to ensure the safe development of AI. Safety here refers to not endangering humanity.

The public admission for its termination of Altman, was him not “being consistently candid” with the Directors. What has also emerged is that under development was an AI, ‘Q Star’, that reportedly, based on internal communication to the Board, was centred on algorithms that presented a threat to humanity.

Whatever the significant dynamics in this event may be, clearly the investors and financial interests scored over any other priority. Possibly the purpose of the Board was either negated or circumscribed. Moody’s outlook has predicted a global enterprise spending of up to 150 billion $, by 2027 on Generative AI.

The approach to reach levels of Human Functionality, by what is termed Gen AI, is galloping. We are cajoled into thinking that it will be ethical and safe. When big money is pushing the frontiers and profit is compelling, the history of Big Business and ethics is desultory.

Also, we now have a new definitional frontier. Artificial General Intelligence is broadly defined as, engaged in developing autonomous systems that are capable of completing most economically valuable tasks, better than humans.

AI is ubiquitous in the media and extended cyber world already. Its role in perception creation and management is pervasive. Understandably there is a flurry of concerns and a core of conscience that seeks its regulation. AI rides the back of the Information, Computer and Technological (ICT) Revolution.

The word ‘revolution’ is a misnomer. Both the domains of ICT and AI are rapidly moving like oceanic currents. Their nature defies classification as well as ring-fencing. How do you regulate constantly moving and mutating domains that develop at the speed of human thought?

AI’s contribution to weapon systems, especially autonomous weapons, is well known. It has enabled a vicious form of no-contact war. But what of its gifts? It has the potential to transform agriculture, industry, education, marketing as well as healthcare to name a few domains. It could play a pivotal role in monitoring and alleviating environmental threats to the Planet.

Unfortunately, some imperatives need to be considered. There is every possibility of AI contributing to a sharpening of the extant enormous societal inequities. It seems poised to address the requirements of the minuscule super-affluent, and those on the fringes of this segment.

It will undoubtedly feed off what Amitabh Ghosh so eloquently states as the “Forever discontented lifestyle of contemporary affluence”. AI will be a gift in medical applications. But what of those who have the finances to pursue immortality? How balanced will the healthcare AI domain be? How ethical will healthcare regulation be, so as not to deny the millions who can’t afford the cutting edge?

AI will undoubtedly benefit manufacturing and industry, as it already does, but will it not result in jobless growth in favour of enormously enhanced profits? A trend that will have disturbing socio-economic consequences.

A precedent has already been seen by large layoffs in IT Companies. Various levels of AI have been used in monitoring climate change and other existential threats, but very few outcomes have been achieved by Nations.

Will AI help the deleterious state of world suffering? In 1994 the number of refugees was 15 million, with an additional five million displaced persons, and a 28 million global population of concern for UNHCR.

In comparison, in mid-2023, there were 36.5 million refugees, over 110 million forcibly displaced persons, and an estimated 300 million that will need humanitarian assistance in 2024.

Cutting-edge AI seems to be leaning in directions favouring profit, weapons and control systems. It could be saddled and bridled to work on gifts for humanity. Left to its own dynamics it will gallop to unknown frontiers and present humanity with unintended consequences.

Lt Gen Sanjiv Langer PVSM, AVSM, is a former Member Armed Forces Tribunal, and former Deputy Chief Integrated Defence Staff. Views expressed here are the writer's own.