Marc van der Chijs redirects AI profits into bitcoin
The Hut 8 co-founder warned that uncontrolled corporate competition creates systemic hazards for banks while forecasting heavy automation across the workforce.
Marc van der Chijs entered the bitcoin market in 2013 and helped establish Hut 8 as a commercial mining operator.12 In recent months, his attention followed the company into artificial intelligence data centers, prompting him to liquidate a large share of his cryptocurrency holdings to finance positions in computational infrastructure.13 That trade delivered substantial returns, yet the velocity of technical change has altered his assessment of where the industry is heading.
A commercial race among leading technology developers and national governments has eliminated incentives for caution, leaving advanced systems to develop faster than institutional safeguards can follow.14 Speaking on his personal outlook, van der Chijs said his perspective on machine intelligence turned sharply more pessimistic over the course of a single week.15
“I’ve become more of a doomer over the past week, to be honest,” he said.16
The competitive trap and systemic fragility
The core of the problem rests on game theory between commercial competitors and rival states. When every firm and government operates under the assumption that restraint guarantees defeat, unilateral pauses become impossible to enforce. Van der Chijs compared the technological and market trajectory of artificial intelligence to the early expansion of digital currency, but drew a stark distinction regarding the degree of human oversight remaining in the system.17
“We have lost control, actually,” he said. “And until we get the control back, I’m actually worried that we’re moving too fast.”13
The argument mirrors warnings issued by Dario Amodei, the chief executive of artificial intelligence research firm Anthropic.12 Amodei urged technology leaders to restrain the speed of frontier model development, noting that systems capable of recursive self-improvement risk escaping human governance and inflicting widespread physical and digital damage on basic infrastructure.14 In both assessments, the competitive dynamic ensures that participants treat prudence as a commercial penalty.
Van der Chijs argued that the resulting vulnerabilities could strike the global financial architecture before political leaders muster the will to intervene.15 Many traditional commercial lenders rely on aging, interdependent software stacks that have accumulated decades of operational technical debt. Advanced artificial intelligence models applied across automated environments could expose structural flaws in legacy banking systems, triggering sudden crises of confidence among interconnected institutions.16
The cryptocurrency sector carries parallel exposures, though van der Chijs drew a line between underlying networks and the commercial infrastructure operating on top of them. Off-chain centralized exchanges, custodial platforms, and lending intermediaries face acute software and operational risks.13 The decentralized base protocol of bitcoin itself remains comparatively resilient against automated systemic shocks.
Automation and public revenue
These hazards have not diminished van der Chijs’s expectations for the scale of economic disruption ahead. He projected that artificial intelligence and advanced robotics could eventually replace between 90% and 95% of existing human employment.12 A contraction of that magnitude would reduce the production costs of basic goods and services to nominal levels, but it would dismantle the conventional tax base of industrialized nations.
Modern state budgets rely heavily on payroll deductions and personal income levies to fund public operations, healthcare programs, and physical infrastructure. If automated systems eliminate the majority of human payrolls, sovereign finance would require entirely different revenue mechanisms to avoid fiscal collapse. Van der Chijs suggested that governments might establish statutory taxes on deployed physical robots or implement usage levies on model tokens.14
Enforcing token taxation presents severe operational hurdles. While large-scale centralized application programming interfaces could be monitored and taxed at the server level, open-weight models executed locally on private hardware cannot be tracked through standard regulatory reporting.17 The decentralized dissemination of software creates immediate enforcement friction for any fiscal authority seeking to replace human income taxes with computational tariffs.
Capital allocation between chips and coins
The friction between technological potential and systemic fragility is visible across van der Chijs’s personal portfolio. Having previously divested substantial digital asset holdings to establish exposure to machine intelligence, he concluded that institutional capital flows into computational hardware depressed cryptocurrency markets throughout the preceding cycle. Institutional allocations that many market participants assumed would flow directly into spot digital currencies were captured instead by the construction of specialized compute facilities.
That reallocation prevented bitcoin from reaching the $200,000 to $250,000 price targets that van der Chijs and other industry forecasters had anticipated.15 Having captured profits from the rapid appreciation of machine intelligence infrastructure, he began shifting capital back into digital currency, utilizing regulated exchange-traded funds rather than holding coins directly on commercial trading desks.16
“I sold a lot of my bitcoin. I went into AI,” he said.14
His capital rotation does not diminish his conviction regarding the underlying corporate economics of computational facilities. Hut 8, which traded on public equity markets under the ticker HUT as an energy-intensive proof-of-work miner, pivoted significant power capacity toward artificial intelligence processing.12 Speaking in a personal capacity, van der Chijs endorsed that corporate reorganization, describing the transition into high-performance computing facilities as the best operational decision in the company’s corporate history.13
Were he directing executive operations at the firm today, van der Chijs said he would allocate corporate capital and electrical capacity entirely toward high-performance computing data centers.15 The recurring revenue contracts offered by corporate compute clients provide stronger business predictability than the volatile unit economics of proof-of-work block rewards.
Corporate valuations and the sovereign asset
Commitment to computational infrastructure does not translate into indiscriminate equity buying across leading model developers. Van der Chijs argued that market valuations for pure-play frontier research labs have become stretched relative to their defensible commercial moats. Over a five-year investment horizon, he expects diversified engineering and manufacturing enterprises such as Tesla and SpaceX to outpace Anthropic in enterprise value generation.17
While specialized model builders may eventually expand to become the largest commercial entities in the global economy, their intermediate margins remain exposed to compute costs and competitive open-source replication.16 Integrated enterprises with proprietary robotics, orbital communication networks, and localized manufacturing scale maintain greater resilience against margin erosion.
Yet across financial shocks, fiscal disruption, and competitive corporate reshuffling, van der Chijs reserves a unique structural role for the asset that established his initial investment career. In an environment characterized by systemic vulnerability in commercial banking software and the rapid devaluation of cognitive labor, absolute cryptographic scarcity functions as a durable store of purchasing power.
“I think of all the assets you can hold, if there’s just one, it would be bitcoin.”12
What this rests on
26 sentences trace to 7 sources.
-
1
C
CoinDesk Document · coindesk.com · 19 Sep 2026 Hut 8 co-founder warns of AI risks as he moves profits back into bitcoin (BTC) See the source
-
2
C
-
3
C
-
4
C
-
5
C
-
6
C
-
7
C