- BlackRock flags years or more of lag between AI spending and broader economic benefits
- Energy capacity and other headwinds could slow build-out of AI infrastructure
The rise of generative artificial intelligence (AI) could be a fundamental technological shift, but for investors the question is when all that promise hits the bottom lines of companies not called Nvidia (US:NVDA).
BlackRock’s Investment Institute (BII) says the next phase of development, phase two, which it calls the broadening of AI investment, will take time once phase one, the current ‘buildout’ stage, is over.
“[GDP growth] gains can only come after AI capabilities are fully deployed,” said Simona Paravani-Mellinghoff, co-head of multi-asset strategies and solutions at BlackRock. “That could take years, followed by the typical lag. It took nearly a century for the steam engine to boost productivity… and it was decades before the 1970s-80s computer and tech revolution paid off.”
The stock market winners of the buildout phase have been the suppliers of AI services and key inputs such as microchips, energy, utilities, materials and specialist real estate.
Risks as phase one matures include policymakers possibly putting the brakes on AI to prioritise the power needs of homes and industry. There will also be competition with energy and grid companies (not to mention the Chinese Communist Party) for essential metals and minerals.
Last week, BP’s (BP.) energy outlook said the power needs of data centres that power generative AI look set to “increase electricity demand materially in some markets in the coming years”. Goldman Sachs estimates a single search using large language model (LLM) ChatGPT requires 10 times the electricity of a Google search.
Former vice-president of energy for Microsoft (US:MSFT) Brian Janous said in a recent Goldman Sachs report that data centre demand could lead to a “significant power crunch” in the US. “Over the medium-to-longer term the major constraint [to generative AI development] will be power,” he added.
Phases two and three
BlackRock’s phase two is the broadening of AI investment. It sees the likes of healthcare, financial and communications services companies being the first to implement new ways of working. The risk is that these adaptations create demand for scarce resources before any supply-side or productivity benefits arise, something BlackRock says markets and central banks don’t seem to appreciate yet.
The upshot is that the first two phases may be costly and inflationary ahead of the full payback being realised.
Phase three – the productivity boom – is very uncertain. An enormous question mark hovers around the length of lag time from capital outlay to hyper-efficient nirvana.
BIackRock’s AI experts say it could eventually boost economies and ease inflationary pressure, but leans towards the lower end of estimates that annual US growth will be 0.1 to 1.5 percentage points once the technology’s use is widespread. Goldman Sachs’ senior global economist, Joseph Briggs, puts the cumulative growth from AI in the US for the coming decade at 6.1 per cent, thanks to both “labour reallocation and new task creation” improving productivity.
The 1970s-2010s IT and communications revolution cumulatively has added around 35 per cent to US GDP, for comparison.
There is a final point of nuance, however – the bursts in productivity gains for IT tended to coincide with benign backdrops: financial de-regulation in the 1980s, the post-Cold War peace dividend and globalisation in the 1990s and spells when capital markets were liquid and investment was high generally.
Goldman Sachs’ head of global equity research, Jim Covello, also argued that the internet boom was the opposite to AI. “Truly life-changing inventions like the internet enabled low-cost solutions to disrupt high-cost solutions even in its infancy, unlike costly AI tech today,” he said.
Nowadays, in a world of economic nationalism, high debt and demographic decline there is plenty to hinder productivity gains from AI. Scarcity is another crucial factor, not least when it comes to energy base loads and capacity. BlackRock did conclude on a positive note, for investors anyway: “Even with this uncertainty, the potential for big future rewards supports the case for investing now.”
