AI now represents the biggest bet in history
Last week Walmart, America’s largest retailer, took us a step closer to a world in which artificial intelligence drives consumer behaviour. The company, which has more employees than the NHS and annual revenue that if it were a country would rank it the 23rd largest in the world by income, announced an agreement with OpenAI, the company that owns ChatGPT, the AI chatbot that summarises books, solves equations and writes undergraduate essays.
Under the deal, customers will be able to buy products from Walmart during conversations with ChatGPT, without leaving the AI platform. Imagine asking the omniscient digital interlocutor what is the best type of refrigerator in a certain price range. Its response will ensure that, without you having to go to the retailer’s website, Walmart will send you the model and charge your credit card or digital wallet several hundred dollars; all with one click.
Half of Americans shop at Walmart at least once a week, so the reach of the deal is transformative. It was another coup for OpenAI, the world’s most valuable startup. With a blizzard of deals and product launches this year the company is now emerging as the most dynamic of America’s tech firms. Just this week it debuted an internet browser to rival Google Chrome. The Walmart deal is a first of its kind in giving OpenAI access to customer data, the digital gold of consumer business. Perhaps not least significant, the announcement pushed Walmart’s share price up 5 per cent in a few hours.
Most importantly, the spectacle of the company that remains the cornerstone of America’s huge consumer economy accelerating the adoption of AI is a reminder how dependent the success of the modern US is on this technology.
It’s no exaggeration to say that America’s vast financial wealth, the current strength of its economy, perhaps even its future fiscal health and its strategic muscle are increasingly supported on a large bet that AI will be the most transformative innovation since the industrial revolution.
The scale of AI investment in the US is not just huge, it has been crucial to US economic growth in the past few years. Total corporate spending on AI infrastructure — data centres, power plants to supply them and software development — grew by 40 per cent between 2021 and early this year.
According to JP Morgan, America’s largest bank, AI-related companies have accounted for 75 per cent of stock returns of the S&P 500, 80 per cent of earnings growth and 90 per cent of capital spending. The five big tech companies that are the most heavily invested in AI — Apple, Nvidia, Microsoft, Alphabet (Google) and Amazon — now account for a third of the value of America’s top 500 public companies.
Jason Furman, an economist at Harvard who served as White House chief economist under Barack Obama, estimates that AI investment accounted for almost all the growth in the US in the first half of this year — without it the economy would have expanded at an annual rate of just 0.1 per cent.
OpenAI reported a valuation of $500bn, yet it makes no profit
The AI boom has big political implications. It may be the reason Donald Trump’s tariffs have not seemed to hit the economy as hard as was generally expected. It may also be offsetting the decline in employment growth that has probably been caused by the administration’s crackdown on immigration.
What’s more, if the optimism around AI is justified, it will be a panacea for a host of America’s economic, social and political problems. Nouriel Roubini, the man dubbed Doctor Doom for his repeated forecasts of economic calamity over the past 20 years, now thinks AI will boost productivity so much that the economy could grow at an average rate of 4 per cent a year, rather than the less than 2 per cent we’ve got used to.
That would lift the incomes of American households, many of which have been stagnant for decades. It would transform the US fiscal outlook, generating more than enough tax revenue to take care of the huge deficits the government now runs. And of course it would all underpin US global power, and dramatically enhance its military and strategic capabilities.
Sceptics worry we have been here before. Just about every big new industrial innovation in the past 200 years has generated more optimism — and higher asset prices — than is eventually sustained by the impact it has on economic performance. Railways in the 19th century, the internet boom 25 years ago, were two powerful examples of bubbles that burst when hopes for their long-term growth impact were not ultimately realised. The fear this time is that AI is not just another bubble but one on which the whole US economy has already become dangerously dependent.
The valuations of the companies spending a fortune on AI imply to some a level of future productivity gains that can’t be justified. OpenAI is a case in point. Sam Altman’s company is private so is not required to publish detailed data but it recently reported a valuation of $500 billion. This on revenues this year of an estimated $13 billion — and it makes no profit. For comparison, Walmart has a valuation of $850 billion and makes almost $700 billion in revenue and $20 billion in profits.
No human can say with absolute confidence whether a rapid run-up in the price of an asset or asset class represents a bubble. Occasionally, anticipation of huge returns is justified. And as John Maynard Keynes said, “Markets can remain irrational long after you can remain solvent.”
But since there is evidently so much faith now in the wisdom of this new technology, where better to go for an answer than straight to the source? “Are we in an AI bubble?” I asked ChatGPT this week.
The reply: “Short answer — yes — in several ways.”
No comments:
Post a Comment