The future of artificial intelligence is becoming increasingly uncertain as it shifts from hype to scepticism, along with growing concerns of its use.

Dr Ganesh Natarajan

Artificial Intelligence seems to have crossed the Gartner peak of inflated expectations and is sliding down the trough of disillusionment for many investors and corporate users.

Leaders are concerned that Gen AI is nothing more than a good and speedy summariser of information that can hasten the process of garnering knowledge and forming an opinion, but does not provide value beyond that. And for investors, whose feeding frenzy took Nvidia to a market value over $3 trillion with $2.5 trillion of that increase happening in the past two years and who are investing in OpenAI at an expected valuation north of $150 billion, the threat of the world pivoting away in search of alternate business models to massive processing and large language models is very real. Are we headed for a major crash in AI or is there a slope of enlightened growth to follow?

REDUCING ENERGY COST

A recent issue of The Economist argues that the huge monies flowing into early AI leaders like Nvidia and OpenAI and the $100 billion of angel and VC money poured into AI start-ups may dry up if the present big concerns about energy guzzling by AI are not addressed. The stark reality is that the world cannot afford the energy cost to power the Large Language Models (LLMs) that will be needed in the current process of training AI to be a partner and in some cases a replacement for human effort. The large GPUs and Nvidia chip models of today could also be wrecked by this reality.

There are work arounds to this problem, new AI specialist chip companies may emerge that compete with Nvidia, and LLMs could give way or be supplemented by more narrow AI systems, which choose between reasoning and text generation as their core focus. The computation costs that are today excessive and, in many cases, comparable with LLM training costs, will have to be reduced by a constellation of AI models working in concert rather than the mega LLMs.

Nimble countries like China will see this as an opportunity to find new ways to AI success. In fact, the denial of access to new chip production to other economies by the US may hasten the process of innovation in China and even countries like India and parts of Europe, which have been AI laggards. They may find it worthwhile to explore new models and provide easier funding conduits to the less aggressive PES that worry about burning their fingers with today’s rampaging hectocorn like Nvidia and OpenAI.

IN THE WRONG HANDS 

Celebrity author Yuval Noah Harari, in his recent talk in New York and his recent book Nexus, has also warned that the propensity of large AI to centralise data and provide instant analysis of  behaviour and intent of every citizen in a country can play into the hand of dictatorial regimes. He makes the point that the much-feared KGB in the Soviet Union and possibly in Vladimir Putin’s Russia and the hated regimes of Mussolini, Stalin and Hitler would have benefited greatly from these tools that could have helped them identify potential dissenters and eliminate them even before the thoughts and actions were fully crystallised in the minds of the perpetrators.

Harari also cautions about rogue algorithms that are the creation of some of the internet giant corporations and mentions Myanmar as a case in point where Facebook algorithms sensed that hate posts against the Rohingyas got more attention and chose to hence propagate those much faster than the healing voices. He argues that an algorithm designed with wrong or inadequate rules will have the potential to drive behaviour and initiate actions that may not have been envisaged by the creators themselves. 

While we exult at the capabilities of tech firms to push the frontiers of AI and look forward to the era of Artificial General Intelligence (AGI) and Artificial Super Intelligence (ASI), we need to be wary of both the environmental consequences and threat of AI moving the human out of the loop and taking over the world.

Let us be conscious about how we push our businesses and our lives towards a future filled with AI agents. There is no doubt that we live in interesting times!

The author is chairman of 5F World, GTT Data Solutions and Honeywell Automation India Limited. Views expressed are personal.