Arthur Hayes warns AI infrastructure boom may end in overcapacity, crypto gains

Arthur Hayes warns AI infrastructure boom may end in overcapacity, crypto gains
AI boom risks crypto surge

Heavy spending on artificial intelligence infrastructure is raising fresh debate over whether the sector is building more capacity than future demand can profitably absorb. Arthur Hayes argues the cycle is likely to follow past technology booms, with excess investment leading to a crash and eventual bailout that could lift bitcoin and other crypto assets.

Highlights

  • Arthur Hayes warned at the Gamma Prime Investing Conference that the multi-trillion dollar AI data center boom may result in significant overcapacity and a market correction by late 2027 or 2028.
  • Hayes noted that current AI demand is driven by end users like SpaceX, OpenAI, and Anthropic—none of which are profitable—while suppliers such as Nvidia and memory chipmakers already see earnings, raising questions about future valuation multiples.
  • Rather than shorting AI stocks, Hayes suggests investors consider that any future bailout and excess liquidity could benefit crypto assets, with his Flop project set to launch a tokenized spot market for compute in Q1 2027.

Hayes links AI buildout to later market correction

As reported by CNBC, Hayes said at the Gamma Prime Investing Conference in Singapore that humanity is "wasting multi-trillion dollars" on AI data centers as companies rush to secure computing power for increasingly advanced models.

Hayes, co-founder and chief investment officer of crypto investment firm Maelstrom, said that pattern mirrors earlier large-scale technology rollouts, which in his view tend to be overbuilt before ending in a crash and bailout. He said the current wave of construction will make computing power "extremely cheap and extremely plentiful," creating the conditions for overcapacity rather than durable scarcity.

He said the strain may become visible in late 2027 or 2028, when much of the data center capacity now under construction is delivered and infrastructure providers seek payment from AI companies that have committed to that compute. Hayes named SpaceX, OpenAI and Anthropic among the end users driving demand, while arguing that none of them currently makes money.

At the same time, he said the bullish case for AI is that the technology becomes useful enough over the next 12 months for demand to grow and AI companies to become profitable. He added that some suppliers to the boom, including memory chipmakers and Nvidia, are already making money, though he questioned whether investors are paying the right multiple for expected future earnings.

Crypto positioning and Flop project strategy

Hayes said he does not favor shorting AI companies or making direct bets on falling prices, calling that approach an unattractive investment opportunity. Instead, he said investors should think about how excess liquidity from a future bailout could flow into digital assets, citing bitcoin and other crypto as likely beneficiaries.

That view also informs Flop, Hayes' AI-agent payments venture, which is expected to launch in the first quarter of 2027. He said cheaper and more abundant computing power should allow AI agents to proliferate, while Flop aims to build a spot market for compute where participants receive Flop tokens for supplying GPUs and carrying out AI inference.

Hayes said there is currently no dedicated payments network for AI agents and argued that a tokenized market could fill that gap. In his view, if software agents can directly convert currency into the computing power they need, they are more likely to use that currency within a broader AI economy.

In our earlier article on rising U.S. electricity demand driven by AI and data centers, we noted forecasts for record power consumption in 2026–2027 as compute needs surge. We also outlined how the generation mix is expected to tilt further toward renewables while coal’s share declines, underscoring how the AI buildout is reshaping energy planning and supply.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.