Although major technology firms such as Google and Amazon boast about the vast sums they're investing in AI systems, IBM's chief executive expresses skepticism about the profitability of these ventures.
TL;DR
- IBM CEO Arvind Krishna questions the profitability of massive AI data center investments by tech giants.
- Data centers require huge energy and investment; $80 billion to build one gigawatt, $8 trillion for 100 gigawatts.
- Krishna estimates a 1% probability of achieving AGI with current technology, despite LLM advancements.
- Major cloud providers are spending billions on AI infrastructure, with Google and Amazon increasing capital expenditure forecasts.
Arvind Krishna, who has been at the helm of the legacy tech company since 2020, said even a simple calculation reveals there is “no way” tech companies’ massive data center investments make sense. This is in part because data centers require huge amounts of energy and investment, Krishna said on the Decoder podcast.
Goldman Sachs estimated indicated earlier this year that the global data center market's overall power consumption reached approximately 55 gigawatts, with a small portion (14%) allocated to AI. With the increasing demand for AI, the power needed by the data center sector might surge to 84 gigawatts by 2027, as projected by Goldman Sachs.
Yet, building out a data center that uses merely one gigawatt costs a fortune—an estimated $80 billion in today’s dollars, according to Krishna. If a single company commits to building out 20 to 30 gigawatts then that would amount to $1.5 trillion in capital expenditures, Krishna said. That’s an investment about equal to Tesla’s current market cap.
Collectively, the major cloud providers might contribute approximately 100 gigawatts, he reckoned, yet this figure necessitates an $8 trillion expenditure, and the profitability required to offset this outlay is substantial.
“It’s my view that there’s no way you’re going to get a return on that because $8 trillion of CapEx means you need roughly $800 billion of profit just to pay for the interest,” he said.
Furthermore, due to technology's swift progress, the processors within your data center might soon be outdated.
“You’ve got to use it all in five years because at that point, you’ve got to throw it away and refill it,” he said.
Krishna mentioned that a driving force behind this surge in funding stems from major technology firms competing to be the initial developers of AGI, which signifies an artificial intelligence capable of equaling or exceeding human cognitive abilities.
However, Krishna states that there's a maximum 1% probability this achievement can be realized with our present-day capabilities, notwithstanding the consistent advancement of extensive language models.
“I think it’s incredibly useful for enterprise. I think it’s going to unlock trillions of dollars of productivity in the enterprise, just to be absolutely clear,” he said. “That said, I think AGI will require more technologies than the current LLM path.”
Concurrently, major cloud providers are continuing their substantial expenditures on AI infrastructure, with projections indicating these outlays will amount to $380 billion within the current year.
During its Q3 financial disclosure, Google's parent company, Alphabet, increased its projected capital expenditures for 2025 to between $91 billion and $93 billion, revising its earlier forecast of $85 billion. The Chief Financial Officer also indicated during the firm's quarterly earnings discussion that a “significant increase” in capital spending is anticipated for the upcoming year, partly attributed to greater infrastructure outlays. Amazon, in its third quarter, similarly raised its capital expenditure estimate its outlook to $125 billion, up from a previous projection of $118 billion.
