Aug 21, 2026 · 54 min
Hayes links crypto’s next rally to liquidity and AI compute
Arthur Hayes on Why AI Agents Will Want to Transact in Units of Compute
The episode connects possible monetary easing with a new crypto use case: autonomous AI agents buying and coordinating computational work.
- 1Hayes argues debt pressures could push the Treasury and Federal Reserve toward more liquidity, supporting Bitcoin and ETH.
- 2He presents Flop as a compute-focused network where AI agents can transact directly and coordinate inference work.
- 3Hayes compares AI infrastructure spending with a leveraged real-estate cycle and questions whether Strategy’s Bitcoin model remains distinctive.
Don't miss
Hayes introduces Flop as a network where autonomous AI agents could transact directly for compute, rather than merely using crypto for human payments.
The brief
Arthur Hayes frames Bitcoin’s rally as a consequence of debt-management pressure: if governments cannot sustain current obligations, he expects liquidity support to follow.
His bullish case extends to ETH and Japan, where yen policy and the Bank of Japan could create another liquidity catalyst alongside U.S. financial support.
The conversation then shifts from macro markets to Flop, Hayes’s proposed network for representing computational work and letting AI agents buy and coordinate compute.
Flop’s design assigns miners and validators different roles in processing inference requests, verifying work, and securing a system built around computation rather than payments.
Hayes closes by weighing Strategy’s Bitcoin financing model against a broader market and explaining why BitMEX’s owners chose to shut the exchange down.
What was said on this episode
29 statements · 15 positive · 9 negative · 5 neutral
Money printing to cover capital misallocation will drive Bitcoin sharply higher.
“And so that's how Bitcoin moons, because they will print money to cover up this capital misallocation.”
Listen at 0:02
Treasury and the Fed respond to rising yields with money-printing measures.
“when it looks like the yields are surging towards that level, they engage in some sort of money printing exercise.”
Listen at 2:28
ETH could quickly exceed $5,000 after breaking through $3,000.
“let's say once we break through 3000 level, I think you're really going to start seeing the train moving on ETH and it could quickly eclipse 5000.”
Listen at 5:44
Hayes’s year-end ETH price target is achievable.
“And I think my year end target is within reach.”
Listen at 5:51
ETH’s performance depends primarily on market positioning, not technology or tokenomics.
“This is all about positioning.”
Listen at 6:36
ETH is beginning a major upward move.
“And so I think this is the beginning of a major move.”
Listen at 6:54
The Japanese yen is substantially undervalued because BOJ rates are too low relative to inflation.
“the yen is the most undervalued currency because the Bank of Japan, the BOJ. has kept rates way, way, way too low versus inflation”
Listen at 8:42
Warsh can authorize unlimited Fed balance-sheet expansion to accommodate Japan.
“And so pretty much whenever Warsh feels like it, he can convene the subcommittee, change all the rules, and essentially allow the Fed to expand its balance sheet in an infinite amount to accommodate the Japanese.”
Listen at 12:51
The Fed will expand liquidity to support Japan’s repatriation of capital.
“And that is going to happen, I think.”
Listen at 13:08
Market distortion through money printing and balance-sheet growth will produce a crypto boom.
“which means printed money, which means balance sheet goes up, which means crypto boom.”
Listen at 13:38
AI investment diverted marginal capital away from crypto.
“It's because AI sucked all the capital.”
Listen at 17:11
Maximum-capacity use causes AI chips to fail in under two years.
“But if you run these chips and sweat them at maximum capacity, they burn out much faster than two years.”
Listen at 17:54
AI data-center real estate is being misclassified as technology rather than ordinary real estate.
“And so I think we've told ourselves a story, real estate equals technology, when real estate is just real estate.”
Listen at 18:28
AI infrastructure investment represents a massive capital misallocation.
“It's a massive capital misallocation.”
Listen at 18:55
AI-related financial stresses will produce a crisis larger than 2000 by mid-2028.
“They will coalesce together by, I think, mid-2028. And we're going to have a problem in terms of a financial crisis larger than what happened in 2000.”
Listen at 21:20
AI agents will favor currencies most directly convertible into compute.
“the agent will use the currency that is most easily converted into the thing that they need to survive, which is floating point operations per second performed on some computer.”
Listen at 26:15
Flop will enable agents to pay for decentralized storage and retrieval of memories.
“I think that the second major use case for flop is to be able to pay to store and retrieve your memories on a decentralized, censorship-resistant network.”
Listen at 28:03
Flop could become the unit of account for AI commerce.
“if we think that this is going to be the unit of account for AI commerce”
Listen at 28:20
Flop validators will judge potentially improper network behavior.
“The validators are tasked with essentially being judges if there is any dodgy behavior.”
Listen at 32:32
AI agents need currencies that let them directly purchase compute when required.
“At the end of the day, I want to be able to buy compute when I need to buy compute.”
Listen at 34:48
Flop plans to distribute about 20% of supply through a ten-year airdrop.
“We have about 20% of the supply estimated in 10 years that we're going to give in an airdrop.”
Listen at 39:08
Flop compute will be cheaper than compute on most other platforms.
“the price of flop per compute is extremely cheap, cheaper than what you're going to find on most other platforms.”
Listen at 41:48
Flop will have ongoing annual inflation of roughly 1–2%.
“So there's constant inflation, something around one to 2%.”
Listen at 42:33
Flop aims to rank among the three most valuable cryptocurrencies within two years.
“if we're not top three most valuable crypto in two years, then our assumptions were not valid.”
Listen at 44:44
The agentic economy will become at least as large as the human economy.
“The eagetic economy will be the same or if not larger than the human one.”
Listen at 45:07
Flop must eventually reduce processing latency to milliseconds or nanoseconds.
“if we really believe in this agentic future, that needs to be milli nanoseconds in terms of how fast this needs to go.”
Listen at 45:51
MicroStrategy will no longer be relevant for institutional Bitcoin exposure.
“I don't think that he's going to be relevant anymore.”
Listen at 49:09
Institutional stock-market demand for crypto exposure will bypass MicroStrategy.
“the institutional bid for crypto exposure through a stock market will not go through MicroStrategy.”
Listen at 50:15
MicroStrategy will be unable to acquire Bitcoin at its former scale because alternatives now exist.
“he's not going to be able to buy as much in a decreed manner because there's not a structural reason why i have to hold microstrategy stock to get bitcoin exposure”
Listen at 50:55
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.
Books & mentions
Some links are affiliate links — PodLume may earn a commission if you buy.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.