AI, a bubble or a revolution in progress

After soaring sharply, the markets have gone into pause mode in recent weeks. In the meantime, OpenAI’s CEO has spoken openly about the risk of a bubble, as the tech giants race to invest in generative AI reaches staggering sums.

Meta, Alphabet, Microsoft and Amazon plan to spend no less than 390 billion dollars on AI this year alone, says Anjali Banstianpillai, Senior Client Portfolio Manager at PAM, with the combined development of cloud and AI. According to Morgan Stanley, investment in data centres will probably reach 2.9 trillion dollars by the end of 2028, and McKinsey is betting on a figure of 6.7 trillion dollars by the end of the decade.

Stunning amounts. The markets were not immune. After the impact of the tariffs and the bamboozling of the markets with a low point on 8 April, the Nasdaq soared by 40% in four months, setting a new record on 13 August, reflecting the enthusiasm for artificial intelligence. Far ahead of the broader S&P 500 index, which limited its gains to just under 30%. Over the last few days, with the latest results published by NVIDIA just in line with expectations, investors seem to want to catch their breath.

However, the boom in artificial intelligence currently seems to be troubling Sam Altman, who launched ChatGPT at the end of 2022. His concern is fuelled by the speed and ease with which some AI start-ups are raising hundreds of millions in funding, and whose prospects are not always clear.

However, OpenAI’s CEO remains confident about the future, performance and long-term benefits of AI despite the current hype, but warns that some valuations are now out of control.

Investors overexcited by AI?

“Are we in a phase where investors as a whole are overexcited about AI? I think so”, he said in a recent interview, but remains convinced that AI is the most extraordinary thing to happen in a very long time.

For Brice Prunas, Thematic Equities Manager, Artificial Intelligence, ODDO BHF AM, OpenAI’s CEO’s intervention is not insignificant, and slyly directed at his competitors, with the aim of influencing their analysis.

“In the field of artificial intelligence,” explains Brice Punas, “every statement must be viewed through the prism of ‘game theory’; in other words, through the strategic interaction between the various market players who are waging a fratricidal war towards super intelligence, and in particular to be the first to gain access to it.

In our view, Sam Altman’s statements are aimed at three targets.

Elon Musk and his strategy, which involves heavy investment in computing power, has enabled him to catch up with Open AI and Sam Altman in just 2.5 years.

Mark Zuckerberg is waging a war for AI talent via Meta, but also through acquisitions paid for at twice their value.

Then there is the world of AI start-ups and therefore AI Venture Capitalism, where valuations can become unreasonable given the quality of the workforce under the guise of “AI”.

Finally, Brice Prunas draws a distinction: “Conversely, AI valuations on the stock market are still reasonable, as they can be expressed in earnings multiples. There are a handful of exceptions to this rule in the case of listed companies, but these could be justified by the uniqueness of the assets concerned.

But Matthieu Belongrade, CFA at DNCA, acknowledges that valuations are high overall and urges caution, with a few exceptions: “The second-quarter earnings season was reassuring both in terms of demand for cloud and infrastructure and the first signs of monetisation of AI and NVIDIA…”

Many analysts continue to believe that the real impact in the medium and long term is in fact underestimated.

Er comparing the current AI boom with the dotcom bubble seems excessive.

Anjali Banstianpillai, Senior Client Portfolio Manager at PAM rejects comparisons between the current AI boom and the dotcom bubble by extending this analysis.

The internet bubble, another story

“At the end of the 90s, in the land of dotcoms, there were many cases of over-indebtedness. Few companies were making a profit. At the moment, most companies have very solid profits and very strong cash flows, and they are financing a large part of this growth through these cash flows. In many ways, it’s a bit different,” explains Anjali Banstianpillai, who adds:

“The technology sector today is much more diversified and resilient than it was at the time of the Internet bubble in 2000. At the time, these companies mainly targeted the professional Internet, and focused on growth rather than profitability. Today, technology companies cover many more market sectors, such as communications services (Alphabet, Meta), manufacturing (Uber), finance (Adyen, Visa), consumer discretionary goods (Amazon) and information technology (Nvidia, Palo Alto).”

Another strong trend is mergers and acquisitions, and sufficient cash is ready to be invested.

Jeremy Gleeson, CIO Tech Equity at AllianzGI, analyses these moments of doubt: “It can be difficult to distinguish between opportunity and fashion. The emergence of new emerging markets (developed through innovation), with strong growth over several years, often involves periods of high and temporary investment with low returns. These moments end up creating uncertainty in the assessment process and therefore risk. This was the case at the beginning of the year, with the DeepSeek moment on the markets. Investors became nervous. Personally, I prefer such reactions to complacency!

Jeremy Gleeson added: “But it seems clear that for companies like Nvidia, the market already has some good things built in. On the other hand, we need to be aware of and take into account the continuing increase in capital expenditure by hyperscalers (providers of cloud computing and data management services) , as well as by sovereign states, which are even stepping up their spending.

For Matthieu Belongrade, CFA DNCA “AI represents a new industrial revolution, but as in previous cycles, there will be winners and losers. We advocate diversified exposure to the AI theme: selectively in hyperscalers, but also in semiconductors, data centres, software and especially cybersecurity, which will be essential to the rise of agentic AI.”

Guy Stear, Head of Developed Markets Research at Amundi Investment Institute : “Stock market gains were not widespread across all market segments. To date, the best performances have been achieved by companies that have taken advantage of AI and those that have participated in the development of this technology (in the form of chips and server farms). We expect this divergence to continue.

Jackson Hole, reassuring

Last but not least, Dr Karsten Junius, Chief Economist, J. Safra Sarasin, provides an insightful and interesting assessment of the Jackson Holes conference, a key economic event.

“For investors, the findings and summary of this meeting are essential and full of lessons to be learned. If inflation can be brought under control without a deep recession, the risk of monetary policy-induced slowdowns diminishes. This justifies higher asset valuations.

Finally, Sam Altman added in his interview, in a mixture of caution and optimism: ” The collapse of the dotcoms led to the demise of many businesses, but nevertheless gave birth to the modern internet. He expects AI to follow a similar path: a few high-profile collapses, followed by a lasting transformation.”

And he concludes: “I think that some investors are likely to suffer heavy losses, and that’s regrettable. I don’t want to play this down. But overall, I’m convinced that… the value created by AI for society will be considerable.”

A must?

It should also be noted that, historically, the great technological revolutions that have brought about a profound transformation of the economy, such as the steam engine at the beginning of the 19th century, the railways in the 1880s, the beginnings of aviation at the beginning of the last century or, more recently, information technology and the Internet in the 2000s, often go through a phase of stock market hysteria and speculation as they mature, followed by a few, sometimes violent, downturns. One excess follows another, but often the story continues..

As ever, the market is not immune to a correction. September is reputed to be the worst month on the stock market, particularly in the second fortnight, but that is just one statistic. However, analysts remain convinced of the future of AI in the medium and long term.