In doing so, he largely echoed what Rick Rieder, his counterpart at giant BlackRock, said not long ago: “If technology has always aimed to do more with less, AI is a breakthrough in doing more with far less. The result? In economic terms: disinflationary growth.”
Why the Fed Should Delay Rate Hikes
However, such statements are not only meant to reassure the public that the world may be approaching another era of low inflation and low interest rates. They also carry a strong lobbying message.
Hunstad added to his prediction a recommendation for the US central bank, arguing that it should refrain from restrictive monetary measures while assessing the impact of the war in Iran and a possible blockade of the Strait of Hormuz on US prices.
He maintains that the Federal Reserve should first assess the impact of new technologies before debating interest-rate changes, as AI could help it reach its 2% inflation target without tightening the money supply.
In a similar vein, Elon Musk has also pushed the idea that the state may one day need to provide people with unconditional income.
He argues that issuing more money by the central bank would normally raise prices, but “if AI/robots massively increase goods & services output, then you actually MUST issue dollars to people or there will be massive disinflation”.
That would not be desirable if it tipped into deflation, as there is broad agreement in mainstream economics that falling prices discourage economic activity. Consumers delay purchases, companies face weaker demand and production slows, creating the risk of a dangerous downward spiral.
AI Rates Its Own Impact More Cautiously
There are, however, different views on how strongly AI will affect labor productivity and whether it will allow central banks to lower interest rates without reigniting inflation. Most economists agree productivity will rise, but they differ on the timeframe and the scale.
An interesting contribution to the debate came from analysts at Germany’s Deutsche Bank, who asked three different models - Google’s AI-powered dbLumina agent, OpenAI’s ChatGPT-5.2 and Anthropic’s Claude Opus 4.6 - for their forecasts.
All three predicted only minimal effects on inflation over a one-year horizon. Paradoxically, they also judged a significant decline in prices to be extremely unlikely and suggested inflation was actually more likely to rise because of AI.
Over a longer five-year horizon, a disinflationary scenario becomes somewhat more likely, but the downward effect on inflation still remains close to zero.
“DbLumina finds it more likely that AI lifts inflation by any amount over the five-year horizon (25%) than meaningfully reduces inflation over that period (20%). That skew is inconsistent with the consensus view of AI as a disinflationary force”, the economists wrote.
The other two models assigned similar probabilities. Over a five-year horizon, all three suggested that the most likely outcome - around 35% on average - was only a marginal reduction in inflation of 0.1 to 0.4 percentage points.
Is the AI Craze a Bubble?
The Deutsche Bank analysts concluded that artificial intelligence is likely to have only a limited effect on inflation because the entire sector is highly capital-intensive and currently absorbing enormous investment flows.
Those investments are driving the construction of data centers and demand for chips, semiconductors, raw materials and electricity. None of that is helping to cool prices in the short term.
This point was also made by Federal Reserve Vice Chair Philip Jefferson, who noted that the inflationary effects of demand growth are immediate, while productivity gains emerge only later and remain uncertain in scale.
Some analysts in the investment community go even further. They argue that technology companies are effectively circulating money among themselves in a vicious circle, which in turn is creating the observed investment boom.
The mechanism was recently described by Bloomberg. Nvidia, which develops the most advanced chips and is currently the world’s most valuable company, has invested $30bn in OpenAI. OpenAI, in turn, plans to equip its data centers with Nvidia chips. Earlier reports had suggested the figure could reach $100bn.
A similar pattern applies in cloud and web services. Companies are buying one another’s products with money they have borrowed from one another. Besides Nvidia and OpenAI, the financially intertwined network also includes Intel, AMD, Microsoft, Oracle and CoreWeave.
The question is not so much whether such a bubble exists, but how large it really is. The same applies to artificial intelligence itself: the question is not whether it increases productivity, but how significantly.
That matters especially in the physical world, where generating outputs is not enough. Someone must still be accountable for them, control them and ensure they can be implemented effectively.
The answers to these questions will determine whether high inflation really becomes easier to fight in the future. They will also show whether, as Fed Chair nominee Kevin Warsh argues, advances in artificial intelligence will truly allow central banks to cut interest rates without risking another rise in inflation.
