AI faces US$6 trillion test to justify data centres: report

[BENGALURU] The global artificial intelligence industry needs to earn US$6 trillion in annual revenue by 2031 to justify the capital being deployed to build data centres around the world, Bain & Company said.

Existing consumer and enterprise AI services may generate as much as US$1.8 trillion of that sum, leaving US$4.2 trillion in new revenue that needs to be created, the consulting firm said on Tuesday (Sep 29) in its annual global technology report.

The shortfall likely will come from nascent segments ranging from autonomous machines to robotics as well as emerging fields such as drug discovery, mental health and energy generation, it said.

“What the industry needs is a wave of innovation that will dwarf what mobile and cloud unlocked,” said David Crawford, the report’s lead author and chairman of Bain’s global technology, media, and telecommunications practice.

“AI infrastructure is being built well ahead of the demand curve and funding it sustainably will require adding approximately 1 per cent to the annual global gross domestic product growth rate,” he said.

Bain’s report highlights the hurdles ahead to make the current pace of AI development sustainable.

Companies led by Microsoft, Alphabet’s Google, Amazon, Meta Platforms and Oracle are investing trillions of US dollars in data centres to quench AI’s demand for computational resources.

Data centre sizes and costs are doubling roughly every 12 to 16 months, in part due to surging prices of chips from the likes of Nvidia and SK Hynix, networking equipment and other components.

The report comes as debate intensifies around the as-yet elusive returns for AI service providers.

SEE ALSO

‘How many will survive?’: Bubble fears arise as China’s humanoid robotics face reality check

Vietnam’s data-centre investment rush faces reality check on power and site constraints

Singapore to prioritise high-value AI infrastructure as Asean markets play to their strengths: Tan Kiat How

Critics worry about an increasingly interconnected web of dependencies between technology manufacturers and AI developers that help to propel lofty expectations that in turn require bigger sums of money.

While discussions today are fixated on employee productivity, the economics of AI infrastructure will demand trillions in new revenue beyond productivity gains, Bain said.

The consultancy projects US$5 trillion to US$6.5 trillion of data centre spending by 2030, adding at least 150 gigawatts of capacity that will further strain countries’ energy resources.

Annual spending on AI infrastructure – spanning data centres, computing capacity and upgrades in accelerators and memory chips – may reach as much as US$1.5 trillion by 2031, it said.

Data centre developers already face shortages in transformers and water and power supplies, as well as fierce local opposition that has blocked or delayed US$68 billion worth of projects in the June quarter in the US. BLOOMBERG