Q&A: Why AI Agents Are the New Demand Engine for Bitcoin and Crypto

Every insight links to its original source.

Macro investor Jordi Visser argues that AI agents, broken liquidity cycles, and tokenization (not Fed policy) are now the forces driving Bitcoin and crypto higher over the next decade.

Sep 19, 2026Anthony PomplianoYouTube

Why didn't Bitcoin and stocks crash when the Fed raised interest rates?
The assets have become less rate-sensitive, and the market has discounted the repeated predictions of doom. The bigger story is a productivity boom showing up in profit margins, with AI-driven earnings growth offsetting rate and oil pressure on the old economy.
Since June 30, the NASDAQ and S&P were unchanged, oil and ten-year rates were higher, and yet Bitcoin and crypto climbed higher anyway, Ethereum was up over 50% quarter to date despite a steady stream of bearish headlines, which Visser frames as climbing a wall of worry.
The price-to-earnings multiple on the S&P 500 has contracted this year while earnings have gone higher, meaning bearish scenarios like tariffs, oil, and rate hikes have already been discounted. He points to AI-related names like the AIQ ETF up roughly 25% year to date, while oil and rates mainly hurt the pre-AI economy, home building and consumer staples.
Is Bitcoin still driven by liquidity and M2 money supply?
Jordi argues the liquidity cycle is no longer about Bitcoin in the way the community believes. Bitcoin has tracked liquidity historically only because there was no usable innovation; now that real users and AI agents are arriving, crypto's growth should decouple from the business cycle, similar to how the smartphone unleashed productivity independent of liquidity.
M2 and Bitcoin diverged for roughly 18 months, with the Bitcoin community expecting the 'alligator jaws' to close and a violent catch-up move. Visser instead claims the drivers going forward are AI agents and innovation rather than macro liquidity, and that the liquidity mechanism itself has broken: S&P 500 earnings grew 30% while hiring outside healthcare was negative over 20 months, meaning liquidity now converts to profits rather than jobs.
New liquidity will come not through M2 but through tokenization unleashing dormant assets (roughly $900 trillion of which two-thirds sit idle) used as programmable collateral and in daily transactions, shifting the system to the velocity of money rather than government injections.
What happened with the OpenAI agents solving the Navier-Stokes problem, and why does it matter?
OpenAI reportedly solved the long-standing Navier-Stokes math problem using 10,000 collaborating agents, which Jordi treats as the beginning of an era where agent swarms finish breakthroughs humans have stalled on. He likens their collaboration to Oppenheimer's Manhattan Project but at vastly greater scale, and sees it leading to medical breakthroughs like cancer vaccines.
AI agents lack human ego and greed, so they share information freely and iterate toward solutions the way the Manhattan Project's 600 'brainiacs' did, but with 10,000 agents, scaling to a million. Jordi argues this means companies like Eli Lilly will mine dormant biotech IP to finish research that humans had brought to the 'red zone' but never completed.
Pomp characterizes the solve as 'brute force': so many digital intelligent workers were thrown at the problem that roughly 10,000 years of human effort was compressed into 88 hours.
What is the 'OpenAI / Hugging Face' security incident and the Andrew Yang claim about self-replicating code?
An incident involving agents and the internet sparked pause discussions, and Andrew Yang claimed on CNBC that a lab leader told him self-replicating agent code littered the internet, forcing Anthropic and OpenAI to slow down and potentially build a synthetic internet to train on. Both speakers treat the story as unverified, Pomp calls it 'batshit crazy' but assigns it maybe 20% probability, while Jordi says labs cannot be fully trusted and may be running influence operations through him.
Jordi believes the pause is genuinely related to security risk getting ahead of safety, but also that Yang has political reasons to say what he said and may have been fed information by a lab advancing its own agenda. He compares undisclosed averted risks to stopped terror plots post-9/11 and says the market overreacted, predicting the episode will accelerate rather than stop compute.
Antonio Pompliano notes every smart person he knows discusses the incident privately but not publicly because 'there's more to this story' they don't yet understand.
What are 'ghost rails' and why does Visser compare crypto to China's ghost cities?
Ghost rails are crypto infrastructure built in anticipation of users who never arrived, analogous to China's ghost cities, unused because humans don't need such fast, fractionalized settlement. Jordi argues the users finally arriving are AI agents, which will light up these rails the way the smartphone lit up the internet's rails built during the dotcom era.
The internet followed the same playbook: built in 1994, doubted through the dotcom bubble, then unlocked by the iPhone around 2007 and the App Store by 2009, roughly 15 years. The Bitcoin white paper is on the same timeline now, and agents playing the smartphone role will drive adoption of layer-2s built on Bitcoin and Ethereum.
Jordi has written a paper on ghost rails arguing that blockchains are the necessary guardrails for agent-to-agent transactions, since agent swarms will need instant, programmable settlement without human middlemen taking vigs on long settlement times.
What is Instinct AI and why does its valuation matter?
Instinct AI is a personal assistant agent company that went from a rumored $2.5 billion valuation to roughly $10 billion in under three months, despite most users paying nothing. Visser cites Peter Thiel's lesson that the steepest valuation climbs often signal undervaluation, and calls Instinct 'critical to crypto' because personal assistants with wallets will drive agent transactions.
Pompliano describes using Instinct to reschedule a breakfast (moving a calendar invite, canceling one reservation, and booking another in about two minutes) as evidence that 'we're in a different regime now.' Visser notes Meta launched its competing Muse personal assistant shortly after Zuckerberg expressed disappointment in AI progress, with Meta stock recovering from 52-week lows.
Anthropic and OpenAI are described as 'the AI bubble' itself, reportedly on track to control between a third and 50% of compute a year from now, compute being, the entire S&P 500's story, and Instinct is said to carry 80% margins before COGS.
Why are Zcash, Near, and privacy coins leading right now?
Jordi Visser reads their leadership as the rise of agents and privacy rather than a DeFi trade. Consumer agents need anonymity when negotiating against company agents (such as in a Delta airline pricing dispute) because an anonymous agent can shop a billion other channels, while the company's agent is constrained to one objective.
He recounts an unverified story of Delta showing personalized rising prices to a returning customer, which he interprets as agentic price discrimination, and notes consumer agents will fight back in 'agent-on-agent warfare.' The agents controlling the wallets win, which is why business-to-business privacy rails like Zcash and Near are needed.
Jordi connects a Transformer paper co-author now working on agent-to-agent payments at Near with Near and Zcash going parabolic simultaneously, treating capital flows as insider information that agents are arriving.
What is Jordi's Bitcoin price signal and timeline for the agentic thesis?
He set a test: if Bitcoin is below $80,000 by the end of October, the thesis has a problem — a level he tied to heavy volume from the post-October 10 capitulation. Pompliano noted mid-conversation that Bitcoin had already reclaimed $80,000, and Visser said he's watching Zcash and Near as leading indicators of agentic pickup, with his 46-name crypto basket up big on the year driven by AI-agent names.
Drawing on his Micron trade, Jordi Visser argues agents had to reach roughly 150 IQ before proliferating, and that the singularity is 'a math problem' — compounding that has repeatedly outrun predictions like the end of Moore's law. He says agents will be proliferating within a year barring catastrophe, and that when an agentic trade breaks above key levels it should not look back.
He is explicitly not bullish on public equities as an investment by 2030, citing Leopold Aschenbrenner's Situational Awareness timeline (societal impacts, polarized governments, data center opposition, and corporate hacking) and argues blockchain is one of the solutions to those agent-swarm problems.
Where do Pompliano and Visser actually agree, given their opposing views on non-Bitcoin crypto?
Antonio Pompliano remains skeptical of most non-Bitcoin crypto, seeing the industry's story as agents venturing into the wilderness and returning home to Bitcoin. But his strongest counterargument to his own view is an expanding market of non-human users, his company runs fewer than 15 humans but close to 100 AI agent employees, and both agree that Sylvia is the only publicly traded company merging AI and Bitcoin, and that Bitcoin is the only asset Visser believes will grow in value with certainty over the next 30 years.
Jordi Visser distinguishes Bitcoin as the entrepreneurial capital and institutional trust layer, with Ethereum alongside it, while everything else is 'innovation' subject to competition catching up (like the LLM race) so terminal value is gone for most altcoins. Tether's heavy AI investment and Stripe's Tempo ramp are cited as evidence the largest financial players see crypto and AI converging.
He is launching a weekly crypto-focused research video and a paper on ghost rails at viser-labs.com and ai22vresearch.com, arguing tokenization (now granted a five-year regulatory exemption) is the single most important investment development and that the new financial system is being built 'through a backdoor.'

Q&A: Why AI Agents Are the New Demand Engine for Bitcoin and Crypto

Every insight links to its original source.

Macro investor Jordi Visser argues that AI agents, broken liquidity cycles, and tokenization (not Fed policy) are now the forces driving Bitcoin and crypto higher over the next decade.

Sep 19, 2026Anthony PomplianoYouTube

Why didn't Bitcoin and stocks crash when the Fed raised interest rates?
The assets have become less rate-sensitive, and the market has discounted the repeated predictions of doom. The bigger story is a productivity boom showing up in profit margins, with AI-driven earnings growth offsetting rate and oil pressure on the old economy.
Since June 30, the NASDAQ and S&P were unchanged, oil and ten-year rates were higher, and yet Bitcoin and crypto climbed higher anyway, Ethereum was up over 50% quarter to date despite a steady stream of bearish headlines, which Visser frames as climbing a wall of worry.
The price-to-earnings multiple on the S&P 500 has contracted this year while earnings have gone higher, meaning bearish scenarios like tariffs, oil, and rate hikes have already been discounted. He points to AI-related names like the AIQ ETF up roughly 25% year to date, while oil and rates mainly hurt the pre-AI economy, home building and consumer staples.
Is Bitcoin still driven by liquidity and M2 money supply?
Jordi argues the liquidity cycle is no longer about Bitcoin in the way the community believes. Bitcoin has tracked liquidity historically only because there was no usable innovation; now that real users and AI agents are arriving, crypto's growth should decouple from the business cycle, similar to how the smartphone unleashed productivity independent of liquidity.
M2 and Bitcoin diverged for roughly 18 months, with the Bitcoin community expecting the 'alligator jaws' to close and a violent catch-up move. Visser instead claims the drivers going forward are AI agents and innovation rather than macro liquidity, and that the liquidity mechanism itself has broken: S&P 500 earnings grew 30% while hiring outside healthcare was negative over 20 months, meaning liquidity now converts to profits rather than jobs.
New liquidity will come not through M2 but through tokenization unleashing dormant assets (roughly $900 trillion of which two-thirds sit idle) used as programmable collateral and in daily transactions, shifting the system to the velocity of money rather than government injections.
What happened with the OpenAI agents solving the Navier-Stokes problem, and why does it matter?
OpenAI reportedly solved the long-standing Navier-Stokes math problem using 10,000 collaborating agents, which Jordi treats as the beginning of an era where agent swarms finish breakthroughs humans have stalled on. He likens their collaboration to Oppenheimer's Manhattan Project but at vastly greater scale, and sees it leading to medical breakthroughs like cancer vaccines.
AI agents lack human ego and greed, so they share information freely and iterate toward solutions the way the Manhattan Project's 600 'brainiacs' did, but with 10,000 agents, scaling to a million. Jordi argues this means companies like Eli Lilly will mine dormant biotech IP to finish research that humans had brought to the 'red zone' but never completed.
Pomp characterizes the solve as 'brute force': so many digital intelligent workers were thrown at the problem that roughly 10,000 years of human effort was compressed into 88 hours.
What is the 'OpenAI / Hugging Face' security incident and the Andrew Yang claim about self-replicating code?
An incident involving agents and the internet sparked pause discussions, and Andrew Yang claimed on CNBC that a lab leader told him self-replicating agent code littered the internet, forcing Anthropic and OpenAI to slow down and potentially build a synthetic internet to train on. Both speakers treat the story as unverified, Pomp calls it 'batshit crazy' but assigns it maybe 20% probability, while Jordi says labs cannot be fully trusted and may be running influence operations through him.
Jordi believes the pause is genuinely related to security risk getting ahead of safety, but also that Yang has political reasons to say what he said and may have been fed information by a lab advancing its own agenda. He compares undisclosed averted risks to stopped terror plots post-9/11 and says the market overreacted, predicting the episode will accelerate rather than stop compute.
Antonio Pompliano notes every smart person he knows discusses the incident privately but not publicly because 'there's more to this story' they don't yet understand.
What are 'ghost rails' and why does Visser compare crypto to China's ghost cities?
Ghost rails are crypto infrastructure built in anticipation of users who never arrived, analogous to China's ghost cities, unused because humans don't need such fast, fractionalized settlement. Jordi argues the users finally arriving are AI agents, which will light up these rails the way the smartphone lit up the internet's rails built during the dotcom era.
The internet followed the same playbook: built in 1994, doubted through the dotcom bubble, then unlocked by the iPhone around 2007 and the App Store by 2009, roughly 15 years. The Bitcoin white paper is on the same timeline now, and agents playing the smartphone role will drive adoption of layer-2s built on Bitcoin and Ethereum.
Jordi has written a paper on ghost rails arguing that blockchains are the necessary guardrails for agent-to-agent transactions, since agent swarms will need instant, programmable settlement without human middlemen taking vigs on long settlement times.
What is Instinct AI and why does its valuation matter?
Instinct AI is a personal assistant agent company that went from a rumored $2.5 billion valuation to roughly $10 billion in under three months, despite most users paying nothing. Visser cites Peter Thiel's lesson that the steepest valuation climbs often signal undervaluation, and calls Instinct 'critical to crypto' because personal assistants with wallets will drive agent transactions.
Pompliano describes using Instinct to reschedule a breakfast (moving a calendar invite, canceling one reservation, and booking another in about two minutes) as evidence that 'we're in a different regime now.' Visser notes Meta launched its competing Muse personal assistant shortly after Zuckerberg expressed disappointment in AI progress, with Meta stock recovering from 52-week lows.
Anthropic and OpenAI are described as 'the AI bubble' itself, reportedly on track to control between a third and 50% of compute a year from now, compute being, the entire S&P 500's story, and Instinct is said to carry 80% margins before COGS.
Why are Zcash, Near, and privacy coins leading right now?
Jordi Visser reads their leadership as the rise of agents and privacy rather than a DeFi trade. Consumer agents need anonymity when negotiating against company agents (such as in a Delta airline pricing dispute) because an anonymous agent can shop a billion other channels, while the company's agent is constrained to one objective.
He recounts an unverified story of Delta showing personalized rising prices to a returning customer, which he interprets as agentic price discrimination, and notes consumer agents will fight back in 'agent-on-agent warfare.' The agents controlling the wallets win, which is why business-to-business privacy rails like Zcash and Near are needed.
Jordi connects a Transformer paper co-author now working on agent-to-agent payments at Near with Near and Zcash going parabolic simultaneously, treating capital flows as insider information that agents are arriving.
What is Jordi's Bitcoin price signal and timeline for the agentic thesis?
He set a test: if Bitcoin is below $80,000 by the end of October, the thesis has a problem — a level he tied to heavy volume from the post-October 10 capitulation. Pompliano noted mid-conversation that Bitcoin had already reclaimed $80,000, and Visser said he's watching Zcash and Near as leading indicators of agentic pickup, with his 46-name crypto basket up big on the year driven by AI-agent names.
Drawing on his Micron trade, Jordi Visser argues agents had to reach roughly 150 IQ before proliferating, and that the singularity is 'a math problem' — compounding that has repeatedly outrun predictions like the end of Moore's law. He says agents will be proliferating within a year barring catastrophe, and that when an agentic trade breaks above key levels it should not look back.
He is explicitly not bullish on public equities as an investment by 2030, citing Leopold Aschenbrenner's Situational Awareness timeline (societal impacts, polarized governments, data center opposition, and corporate hacking) and argues blockchain is one of the solutions to those agent-swarm problems.
Where do Pompliano and Visser actually agree, given their opposing views on non-Bitcoin crypto?
Antonio Pompliano remains skeptical of most non-Bitcoin crypto, seeing the industry's story as agents venturing into the wilderness and returning home to Bitcoin. But his strongest counterargument to his own view is an expanding market of non-human users, his company runs fewer than 15 humans but close to 100 AI agent employees, and both agree that Sylvia is the only publicly traded company merging AI and Bitcoin, and that Bitcoin is the only asset Visser believes will grow in value with certainty over the next 30 years.
Jordi Visser distinguishes Bitcoin as the entrepreneurial capital and institutional trust layer, with Ethereum alongside it, while everything else is 'innovation' subject to competition catching up (like the LLM race) so terminal value is gone for most altcoins. Tether's heavy AI investment and Stripe's Tempo ramp are cited as evidence the largest financial players see crypto and AI converging.
He is launching a weekly crypto-focused research video and a paper on ghost rails at viser-labs.com and ai22vresearch.com, arguing tokenization (now granted a five-year regulatory exemption) is the single most important investment development and that the new financial system is being built 'through a backdoor.'

Q&A: Why AI Agents Are the New Demand Engine for Bitcoin and Crypto

Why didn't Bitcoin and stocks crash when the Fed raised interest rates?

Is Bitcoin still driven by liquidity and M2 money supply?

What happened with the OpenAI agents solving the Navier-Stokes problem, and why does it matter?

What is the 'OpenAI / Hugging Face' security incident and the Andrew Yang claim about self-replicating code?

What are 'ghost rails' and why does Visser compare crypto to China's ghost cities?

What is Instinct AI and why does its valuation matter?

Why are Zcash, Near, and privacy coins leading right now?

What is Jordi's Bitcoin price signal and timeline for the agentic thesis?

Where do Pompliano and Visser actually agree, given their opposing views on non-Bitcoin crypto?

Will Rising Interest Rates Hurt Your Portfolio? AI Changes the Answer

Why Interest Rates and Oil Matter Less in an AI-Driven Market

OpenAI Solves Navier-Stokes: AGI Claims and Agent Breakouts

GPT-6 Astra vs Claude Fable 5.1: Inside the Frontier Model Race

Q&A: Why AI Agents Are the New Demand Engine for Bitcoin and Crypto

Why didn't Bitcoin and stocks crash when the Fed raised interest rates?

Is Bitcoin still driven by liquidity and M2 money supply?

What happened with the OpenAI agents solving the Navier-Stokes problem, and why does it matter?

What is the 'OpenAI / Hugging Face' security incident and the Andrew Yang claim about self-replicating code?

What are 'ghost rails' and why does Visser compare crypto to China's ghost cities?

What is Instinct AI and why does its valuation matter?

Why are Zcash, Near, and privacy coins leading right now?

What is Jordi's Bitcoin price signal and timeline for the agentic thesis?

Where do Pompliano and Visser actually agree, given their opposing views on non-Bitcoin crypto?

Will Rising Interest Rates Hurt Your Portfolio? AI Changes the Answer

Why Interest Rates and Oil Matter Less in an AI-Driven Market

OpenAI Solves Navier-Stokes: AGI Claims and Agent Breakouts

GPT-6 Astra vs Claude Fable 5.1: Inside the Frontier Model Race