Relax, enjoy the dawn of the AI era, and disregard frightening headlines

From 1980s fears over banking computerisation to today's worries about AI, the author argues that technological change inspires anxiety before it ultimately reshapes society for the better

I distinctly recall how in the mid-1980s, when I was in the first decade of my working life in Mumbai (then Bombay), newspaper headlines were dominated by things like: “Bank unions warn: Computerisation to throw thousands out of work”. 

Even a nerd like me, fresh out of IIM Calcutta and trying hard to get our advertising agency to take off, would wonder: “Is the world coming to an end? Is India about to experience a mass revolution that would change its character?” It was loudly proclaimed, in slogans and headlines (with no private television channels or Internet, these mattered back then), that the 80-odd giant textile mills — the mainstay of Bombay’s economy — were losing money thanks to new technology like polyester and nylon textiles replacing cotton. The workers’ salaries were not being paid. For us, the major advertising clients were these very textile mills … if they shut shop, what would the future of our business be? 

These days, I find many of my friends and acquaintances sounding as if they are about to slide into a similar mood over headlines like: “AI in banking: Employees hear roles may shrink as automation speeds up”. And then there are countless editorials warning about the threat of fake news, and the imminent decline of our tech services giants. 

I try pointing out to them that it is because of the computerisation of banks that we no longer have to stand in a 100-metre-long queue to draw cash from our own accounts, that we don’t need to carry bags full of cash as we once had to, and that far from shrinking their workforce, banks have now extended their reach from 10 per cent of India’s population (primarily in urban centres like Mumbai, Kolkata, Chennai, and Delhi) to 85 per cent, covering previously underserved sections in rural India. When I quote such figures, I can notice their faces light up with relief. 

Listening to contemporary anxiety about AI, I have started wondering whether it was a mistake to give the label “artificial intelligence” to a method that more accurately should have been called “large language models”. The term “artificial intelligence” implies that tools like ChatGPT and Google’s Gemini possess agency, intent and reasoning as humans do. But clearly, calling it AI makes it easier to enthuse investors and raise capital. This is not the first time such an investor-oriented naming has happened. In the 1970s, a computer playing chess was deemed to be using AI to do so. In the 1990s, optical character recognition (OCR) was called “AI”. Now, it’s just a standard feature on your phone. 

An even more accurate description would be “high-dimensional pattern matching”. What we call “AI” today may, in 10 years, just be called “superfast autocomplete”. I asked my frequent consultant Gemini: “Why does the label AI matter instead of, for example, ‘high-dimensional pattern matching’ or some statistical term like that?” The answer I got was: 

“The label ‘AI’ matters because it transforms a statistical tool into a mythological entity. If we called it ‘high-dimensional pattern matching,’ we would treat it as a utility — predictable and human-controlled. By labeling it ‘intelligence,’ we attribute agency to the machine. This framing exponentially inflates its economic valuation while simultaneously diffusing accountability: When a statistical tool fails, we blame the user, but when ‘AI’ fails, we blame the ‘black box’. The label successfully rebrands computation as creation.” 

When I read this, I was forced to scratch my head in bewilderment. Gemini was so very truthful, even though I had expected it to give a marketing-oriented reply. 

As is my usual practice, I asked my other pal ChatGPT the exact same question. This time, the answer was: “We call it ‘artificial intelligence’ not because today’s systems think like humans, but because society needs a simple, evocative label for a sprawling scientific project whose ambition is to build machines that exhibit intelligent behaviour. Technical terms describe what the systems do; AI describes what the field aims for.” 

I was stunned by the truthfulness of this answer as well. 

When we reflect a little, it is apparent that the reason we humans are stunned and impressed by the answers that ChatGPT, Gemini, and other similar tools in vogue today give is their human-like conversational format. The tone and phraseology that they use in their answers, or in the clarifications that they seek from us human users, are human-like. I suddenly remembered that Steve Jobs, when he designed the iPhone, did something similar — he introduced “touch” as a method of navigation (swipe, pinch, zoom), inspired by his study of the yogic principle of pratyahara (the withdrawal of senses from clutter into focus). Thanks to this, we users did not have to painfully type in characters for everything on our mobile phones. The touch-based phone became every human being’s constant and inseparable companion. 

So, let’s all relax and enjoy the new era dawning: The AI era! 

The author (ajitb@rediffmail.com) is devoting his life to unravelling the connections between technology and society