'We have lost control': Crypto pioneer warns AI could trigger systemic banking and infrastructure shocks
Marc van der Chijs sold much of his bitcoin to invest in AI. Now, growing concerns about the technology accompany a move to put some profits back into crypto.
Marc van der Chijs, an entrepreneur who co-founded then bitcoin miner Hut 8 (HUT), has become increasingly concerned about the technology to which the company has pivoted: artificial intelligence.
“I’ve become more of a doomer over the past week, to be honest,” he told CoinDesk in an interview.
The concern marks a shift for an investor who compares AI’s potential today with bitcoin when he entered the market in 2013. He still expects the technology to transform the economy, but increasingly questions whether humans can retain control over its development.
“We have lost control, actually,” he said. “And until we get the control back, I’m actually worried that we’re moving too fast.”
Van der Chijs’s shift toward a "doomer" outlook mirrors warnings issued by Anthropic CEO Dario Amodei, who urged tech leaders to slow the pace of frontier AI development. Amodei similarly cautioned that rapid progress—driven by AI models recursively improving themselves—risks exceeding human control and causing widespread infrastructure damage.
Both men point to the same structural trap: an intense competitive race between companies and nation-states that penalizes individual restraint, making systemic disruption or catastrophic failure seem almost inevitable before true international guardrails are established.
Van der Chihs fears a major disruption, potentially affecting financial systems or critical infrastructure, may be needed before governments cooperate.
The concern extends to banks. Van der Chijs believes AI could expose vulnerabilities in legacy banking software, threatening confidence in institutions that depend on interconnected systems. Within crypto, he sees exchanges and other businesses built around bitcoin as more vulnerable than the underlying network.
His concerns have not displaced an expansive view of AI’s economic potential. He predicts AI and robotics could eventually perform 90% to 95% of existing jobs, while driving sharp declines in the cost of goods and services.
In that scenario, governments would need new sources of revenue as employment changes. He suggested taxes on robots or AI token usage, while acknowledging that locally run models would complicate collection.
Van der Chijs believes that investors redirecting capital toward AI helped keep bitcoin below the $200,000 to $250,000 levels he and others had anticipated. He is now allocating some AI profits back into crypto, primarily through exchange-traded funds.
He also sees a stronger business case for AI infrastructure than for bitcoin mining. Speaking in a personal capacity, he described Hut 8’s move into AI as “the best move ever” and said that, if he were running the company today, he would favor allocating entirely to AI data centers.
Still, enthusiasm for the technology does not mean he considers every AI investment attractive at current valuations. Over the next five years, he believes Tesla (TSLA) and SpaceX (SPCX) could outperform Anthropic, despite his expectation that leading AI developers could eventually become the world’s largest companies.
Bitcoin occupies a different place in his thinking: the asset he would choose if restricted to holding just one for the rest of his life.
“I think of all the assets you can hold, if there’s just one, it would be bitcoin.”
As stablecoins move into regulated finance, APAC is becoming a key proving ground. This report maps the region’s rules, use cases, and RLUSD’s role.
Asanat Analysis — Why it matters
Marc van der Chijs's portfolio rebalancing—liquidating crypto exposure to fund AI, then reversing course—reflects broader macro sentiment shifts in the crypto-native investor class. This isn't merely individual conviction pivoting; it signals that even early-stage believers in AI's transformative potential are hedging against systemic risk vectors that AI deployment introduces. The timing matters: infrastructure shocks from AI-driven systems (market microstructure collapse, energy grid stress, financial system instability) are now concrete enough that diversification into crypto's non-correlated assets looks like portfolio insurance rather than speculation.
The 'loss of control' framing references a structural problem: AI systems optimizing financial flows, energy consumption, or resource allocation at scale lack adequate circuit breakers and operate at speeds humans cannot monitor. Crypto's core value proposition—decentralized, censorship-resistant infrastructure—becomes relevant again in this context, not as a speculative bet but as a backstop against centralized system failures. This narrative tracks with rising institutional interest in Bitcoin as a correlation hedge amid AI capex and energy demand shocks.
Historically, crypto adoption surges during periods of institutional concern about fiat-system stability. An AI-driven systemic event (market flash crash, infrastructure cascade failure, power grid strain) would validate the 'digital cash alternative' thesis in ways that regulatory clarity never could—driving flows back into uncorrelated assets precisely when centralized systems face strain.