Opinion | Are We Already In An AI Bubble?

I've been wrestling with a question that many are asking: are we already in an AI bubble? And if we are, what does that mean for India? To get a grip on where we stand, we first need to look at the global picture - because that's where the story of AI's present and future is being written.

Two pieces got me thinking. The first, from The Economist, highlights how the investor frenzy around AI might be cooling off. Some of the biggest AI companies have seen their stock prices dip by as much as 15% in the last month alone. And here's the real kicker: only about 4.8% of American companies currently use AI in their products or services, down from earlier this year. That's a shocking statistic, given all the hype.

Then there's the Goldman Sachs report titled, 'Gen AI: Too Much Spend, Too Little Benefit?'. In it, Jim Covello, the Head of Global Equity Research, points out something we should all be paying attention to: over $1 trillion is expected to be spent on AI infrastructure and development in the coming years. But will it be worth it?

Covello's not convinced. He argues that for AI to justify that level of spending, it has to tackle some really complex problems - ones it's not quite ready to solve. Developing and running AI isn't cheap. Unlike other tech shifts that helped cut costs, AI doesn't seem to be bringing down expenses.

So the big question is - will these massive investments ever pay off?

The High Price Of An AI Query

According to The Economist, training models like OpenAI's GPT-4 consume massive amounts of energy - enough to power 50 American homes for a century. Costs for training these models are skyrocketing: current estimates for the biggest models hit $100 million, with future generations potentially reaching $1 billion or even $10 billion. The operational costs - answering queries, for instance - can also balloon, costing up to $223,000 for tasks like summarising financial reports. This raises serious concerns for investors despite the $95 billion already poured into AI startups this year.

But here's where things get interesting.

Enter Entrepreneurial Optimism

Despite this scepticism, the entrepreneurial spirit around AI refuses to slow down. Just look at Mukesh Bansal, a veteran entrepreneur who recently launched Nurix.ai and secured $27 million in funding. Mukesh is no stranger to building successful ventures, and his leap into AI signals a strong belief in its potential.

This brings me to a broader observation: we might be on the cusp of a new wave of Indian AI startups ready to go global, much like the SaaS boom we witnessed a few years ago. These startups are not just riding a hype train; they aim to solve real-world problems and bring tangible value to businesses and consumers alike.

Moreover, this massive wave of AI tools and generative AI tech is hitting the market, putting powerful capabilities in the hands of everyday users. It's democratising intelligence like never before - people can now tap into tools that were once only accessible to experts or big corporations. So, while billions -maybe even trillions - are being sunk into AI infrastructure, there's never been a better time to explore and use these tools for yourself.

Finding A Middle Ground

So, where does this leave us? On the one hand, we have warnings of an AI bubble fueled by excessive spending with questionable returns. At the same time, we see relentless optimism from entrepreneurs and investors betting big on AI's future.

The truth lies somewhere in between. Yes, there may be overexuberance, but dismissing AI's transformative potential could mean missing out on significant innovations.

Do you believe we're in an AI bubble? Or do you think this is just the beginning of AI's true potential being realised?

(Pankaj Mishra is the AI Editor at NDTV)

Disclaimer: These are the personal opinions of the author