To AI or not to AI, that’s not the question any more — India must plan to ride it

When the world's most valuable companies are all AI powerhouses, government and society must prepare to handle ubiquitous AI in all spheres of life

Ride the new bronco

The world’s largest companies by market capitalisation are tech companies, specifically, tech companies that produce and deploy AI. This reflects the expectation that AI will play an expanding role in our lives, creating new value in both existing activities and in new activity enabled by, and only by, AI.

The rise of AI presents us with new opportunities, new threats and new challenges. Innovating such response spans education, societal change, regulation and societal consensus on shunning use of AI in a way that realises the worst outcomes that sociobiologist E O Wilson feared could come out of the combination of humankind’s palaeolithic emotions, mediaeval institutions and godlike technology.

Nvidia, designer of chips used to produce and run AI, recently became the world’s most valuable company, beating Microsoft and Apple. Daily price variations make them go up and down in the rankings, but these constitute the most valuable trio. Of the seven companies that command a value of $2 tn or more, all save Saudi Aramco are tech giants — Alphabet, Amazon and Meta are the ones not already named.

Of the lot, all save Apple run and enable AI applications. Apple’s stock price went up recently, after it, too, announced some of its own AI offerings.

What does the dominance of tech companies mean for humanity? Does society need to arm itself with anything new either to take advantage of evermore ubiquitous AI, or to guard against possible harm from such prevalence of AI?

Let us appreciate that AI is already active in our daily lives in assorted ways. The options Google throws up the moment you start typing in the search box, the automatic spelling suggestions that materialise in Word or your phone’s messaging app, the quant trading on the stock market, the anti-lock braking system in your car, and, if you’re of a generation and location that admit use of dating apps, the opportunity to swipe right or left — algorithms are an integral part of your existence, some of them capable of learning as they go along, and making changes accordingly to what they do.

AI, in combination with factory floor robots, low-latency, high-throughput communications among robots and with AI, is poised to revolutionise manufacturing. AI can scan through X-ray and MRI images to spot abnormalities far more efficiently than humans can. AI can produce deepfakes and false messaging at scale to interfere with elections. It can produce art to order, summarise reports, write the first draft of a speech or a technical brochure. It can help a coder write 20 times more code than on his own. It can translate a screenplay into a dozen foreign languages. In other words, AI can destroy jobs as well as create altogether novel economic opportunity.

This creates fresh divides in the world. One is between those who can take advantage of AI and those who cannot, and lose out, either to AI or to those who wield AI to their own advantage. Another is between those who can create AI and AI applications, and those who cannot.

Cultures that have a grasp of the number system, and of infinity and the infinitesimal, find themselves at an advantage, compared to cultures whose quantitative cognition is limited to distinguishing one from two, and these from a numerical portmanteau simply understood as ‘many’.

The divide between the AI haves and the AI have-nots is going to be starker than this numerical divide, and it will manifest itself between nations, and within nations. The way to bridge this gap is to empower those who lack the requisite social and cultural capital. That means not just a more effective education system but also a sharp redistribution of social power to eliminate structural hierarchy in society, of the kind represented by the caste system, which cripples human agency and hobbles the ability to confront adversity and overcome it.

AI poses challenges even for those who are educationally and culturally attuned to effectively make use of it. This can come from badly designed and programmed AI, which picks up existing societal prejudices and amplifies them, such as giving lower eligibility scores to applicants who are either women or bear subaltern caste names. They can fall victim to AI acting out the Goebbelsian norm that a lie repeated a thousand times becomes the truth. They could come up against AI put to nefarious use, such as being used to crack passwords, to troll or blackmail individuals online, or to immobilise their cars remotely.

All these contingencies are over and beyond the possibility of AI going rogue, the classical SF nightmare, whose chances of morphing into daytime reality cannot be dismissed outright.

What this means is that AI calls for sound, thought-through regulation, both specific to AI, and in relation to possible misuse of AI in ways that fall foul of norms that are valid outside the realm of AI as well.

Let us also note that the rise of AI is testament to the continuing and irreversible globalisation of human life. AI enmeshes the lives of people across the world — by empowering them, or by disabling them, by feeding on information on people and their lives to train and extend itself, by enhancing the commercial capability of companies in fields unrelated to AI to do business with, for and against people anywhere in the world.

Whether such interaction with AI produces gain or loss depends on local action. India must plan for it.