Will the race to artificial general intelligence (AGI) take us to the land of financial abundance – or will it end 2008-style? Trillions of dollars depend on the answer.
The figures are staggering: an estimated $2.9tn (£2.2tn) is being spent on datacentres, the central nervous system of AI devices; the more than $4 trillion stock market capitalization of Nvidia, a maker of chips that power cutting-edge AI systems; And Mark Zuckerberg’s meta gave a $100 million sign-on bonus to top engineers at OpenAI, the company behind ChatGPT.
These skyrocketing numbers are all supported by investors who expect returns on their trillions. AGI, a theoretical state of AI where systems achieve human-level intelligence across a range of tasks and are able to replace humans in white-collar jobs like accountancy and law, is a cornerstone of this financial promise.
This offers the possibility of computer systems performing profitable functions without the associated costs of human labor – an extremely attractive scenario for the companies developing the technology and the customers who deploy it.
If AI companies lag behind, there will be consequences: The performance of tech stocks could cause the booming US stock market to collapse and people’s personal wealth to suffer; Debt markets wrapped up in the datacenter boom could face a shock that could spread elsewhere; GDP growth in the US, which has benefited from AI infrastructure, could falter, which would have knock-on effects across interconnected economies.
David Kahn, partner at Sequoia Capital, a leading Silicon Valley investment firm, says tech companies now have to work on AGI.
“Anything less than the AGI will not be high enough to justify the proposed investments for the coming decade.” He wrote in a blog published in October.
This means a lot hangs on progress toward advanced AI, and trillions are being spent in infrastructure and R&D to achieve it. Yoshua Bengio, one of the “godfathers” of modern AI, says AGI progress could grind to a halt and the outcome would be bad for investors.
He says, “There is a clear possibility that we will hit a wall, that there is some difficulty that we do not yet anticipate, and that we do not find a solution quickly.” “And that could be a real (financial) crash. Many of the people who are investing trillions in AI right now are also expecting progress to continue at the current pace.”
But Bengio, a leading voice on the security implications of AGI, is clear that continued progress toward a highly advanced state of AI is the more likely end.
He says, “Stopping advances is a minority scenario, just as it is an unlikely scenario. What is more likely is that we will continue to move forward.”
The pessimistic view is that investors are supporting an unrealistic outcome – that AGI will not happen without further breakthroughs.
David Bader, director of the Data Science Institute at the New Jersey Institute of Technology, says trillions of dollars are being spent — tech jargon for developing something faster — on the underlying technology for chatbots, known as Transformers, in the hope that increasing the amount of computing power behind current AI systems by building more datacenters will be enough.
“If AGI requires a fundamentally different approach, perhaps something we haven’t imagined yet, then we’re optimizing an architecture that can’t get us there no matter how big we make it. It’s like trying to reach the moon by building higher stairs,” he says.
Nevertheless, large US tech companies such as Google’s parent company Alphabet, Amazon and Microsoft are moving forward with datacenter plans with the financial comfort of being able to fund their AGI ambitions through cash generated from their highly profitable day-to-day businesses. This at least gives them some protection if the wall outlined by Bengio and Bader comes forward.
But there are more worrying aspects to the boom. Analysts at US investment bank Morgan Stanley estimate $2.9 trillion will be spent on datacenters between now and 2028, half of which will be covered by the cash flows of “hyperscalers” such as Alphabet and Microsoft.
The rest will have to be covered by alternative sources such as private loans, a corner of the shadow banking sector that is raising alarm bells at the Bank of England and elsewhere. Facebook and Instagram owner Meta has borrowed $29 billion in the private credit market to finance a datacenter in Louisiana.
According to investment bank JPMorgan, AI-related sectors account for about 15% of investment grade loans in the US, which is larger than the banking sector.
Oracle, which signed a $300 billion datacenter deal with OpenAI, has increased credit default swaps, a form of insurance against a company defaulting on its loans. High-yield, or “junk debt”, which represents the higher-risk end of the lending market, is also appearing in the AI sector through datacenter operators CoreWeave and TeraWulf. Growth is also being financed by asset-backed securities – a form of debt backed by assets like loans or credit card debt, but in this case rent paid by tech companies to datacenter owners – a form of financing that has grown rapidly in recent years.
It’s no surprise that JPMorgan says the boom in AI infrastructure will require contributions from all corners of the credit market.
Bader says: “If AGI does not materialize on the expected timelines, we could see a transition to multiple debt markets simultaneously – investment-grade bonds, high-yield junk debt, private debt and securitized products – all of which are being used to fund this creation.”
Stock prices related to AI and technology are also playing a big role in the US stock markets. The so-called “Glorious 7” of US tech stocks – Alphabet, Amazon, Apple, Tesla, Meta, Microsoft and Nvidia – account for more than a third of the value of the S&P 500 index, America’s largest stock market index, compared with 20% at the beginning of the decade.
In October the Bank of England warned of “risks of a sharp correction” in US and UK markets due to the volatile valuations of AI-linked tech companies. Central bankers worry that the stock market could crash if AI fails to reach the transformative heights investors are expecting. At the same time the International Monetary Fund said valuations were heading toward dotcom bubble-levels.
Even tech executives whose companies are increasingly profitable are acknowledging the speculative nature of the mania. In November, Alphabet chief executive Sundar Pichai said there were “elements of irrationality” in the boom and that “no company would be untouched” if the bubble burst, while Amazon’s founder, Jeff Bezos, has said the AI industry is in “a kind of industrial bubble”, and Sam Altman, chief executive of OpenAI, has said, “There are many parts of AI that I think are kind of in a bubble right now.” There are bubbles.”
Clearly, all three are AI optimists and hope that the technology will continue to improve and benefit society.
But when the numbers get so big, there is a clear risk of the bubble bursting, as Pichai acknowledges. Pension funds and anyone invested in the stock market will be affected by the fall in share prices, while the debt market will also be affected. There is also a web of “circular” deals, such as OpenAI paying Nvidia in cash for chips, and Nvidia investing in OpenAI for non-controlling shares. If these transactions get settled due to lack of AI or because they hit that wall, it could get messy.
There are also optimists who argue that generative AI, the ubiquitous term for tools like chatbots and video generators, will transform entire industries and justify the expense. Benedict Evans, a technology analyst, says the spending numbers are not outrageous in the context of other industries, such as oil and gas extraction, which runs up to $600 billion a year.
“These AI capex figures are huge amounts, but not insurmountable amounts,” he says.
Evans says: “You don’t have to believe in AGI to believe that generative AI is a big thing. And what’s happening here isn’t, ‘Oh wow, they’re going to create God’. It’s ‘This is going to completely change how advertising, search, software and social networks – and everything else that our business is based on – will work.’ It’s going to be a huge opportunity.”
Nonetheless, the trillion dollar AGI is expected to be achieved. For many experts, the consequences of getting there are worrying. The cost of not getting there can be quite high.
