Hut 8 co-founder flags AI systemic risk and rotates some gains back into bitcoin ETFs
Marc van der Chijs said he sold substantial BTC to fund AI bets and is now reallocating some AI profits back to crypto exposure.
Hut 8 co-founder Marc van der Chijs is pairing a sharper warning that the AI race is slipping beyond human control with a quieter portfolio tell: he is rotating some AI profits back into bitcoin exposure via ETFs. The shift comes after he said he sold a lot of his bitcoin to invest in AI, even as he argues AI data centers are a better business than bitcoin mining.
Hut 8 Co-Founder’s AI “Doomer” Turn Comes With a BTC ETF Rotation
Marc van der Chijs, who co-founded Hut 8 (HUT) when it was primarily a bitcoin miner before its pivot toward artificial intelligence infrastructure, said he has become “more of a doomer over the past week, to be honest,” as competitive pressure among companies and nation-states accelerates frontier AI development past what humans can reliably control.
“We have lost control, actually,” van der Chijs said. “And until we get the control back, I’m actually worried that we’re moving too fast.” He framed the dynamic as a race that punishes restraint, where the incentives to keep building can outrun the incentives to slow down.
For traders, the more actionable detail was not the broad warning but the way he described managing his own book. “I sold a lot of my bitcoin. I went into AI,” he said, adding that he is now allocating some AI profits back into crypto, primarily through exchange-traded funds.
That admission matters because it explicitly treats AI and bitcoin as competing allocations, not just adjacent narratives. Van der Chijs also argued that investor capital redirected toward AI helped keep bitcoin below the “$200,000 to $250,000” levels he and others had anticipated, though he did not provide flow data or positioning evidence to support the claim.
Systemic-Risk Framing: Legacy Bank Software, Interconnected Confidence, and Critical Infrastructure
Van der Chijs’s systemic-risk thesis is bank-centric in its mechanics. He said AI could expose vulnerabilities in legacy banking software, meaning older core systems banks rely on for transactions and record-keeping that can be difficult to secure or update quickly, and that those weaknesses could threaten confidence in institutions that depend on interconnected systems.
That “confidence” channel is the part macro traders tend to map onto risk assets fast, because it is less about a single breach and more about second-order effects: if one institution’s operational integrity is questioned, counterparties and customers can pull back, and the stress propagates through linkages that were designed for efficiency, not resilience.
He extended the risk framing beyond finance to critical infrastructure, arguing that the same competitive race could produce shocks that force governments to coordinate on guardrails only after a major disruption. The interview did not include specific technical examples, dates, or incident data for the banking-software vulnerability claim, so the risk is presented as his assessment rather than a documented failure mode.
Inside crypto, van der Chijs said exchanges and other businesses built around bitcoin are more vulnerable to AI-driven disruption than the underlying bitcoin network itself, without naming specific venues or citing incidents. That distinction is familiar in past market stress, where the plumbing around an asset can fail before the asset’s base-layer rules do.
His view on AI remains economically bullish even as his tone turns darker. He predicted AI and robotics could eventually perform “90% to 95%” of existing jobs and sharply reduce costs, and he floated the idea that governments may look for new revenue sources such as taxes on robots or AI usage, including “AI token usage,” while acknowledging that locally run models would complicate enforcement.
He also drew a clean line between asset conviction and business-model preference. Speaking personally, 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 allocate entirely to AI data centers. At the same time, he said, “I think of all the assets you can hold, if there’s just one, it would be bitcoin.”
Positioning Checklist: What Would Confirm the AI-to-BTC Capital Rotation Narrative
The first confirmation point is basic but missing: which bitcoin ETFs he is using, and whether the reallocation grows from “some AI profits” into a repeatable sizing decision. The interview did not disclose timing, amounts, or tickers, which makes it hard to separate a one-off rebalance from a durable shift.
The fastest way the systemic-risk framing would reprice is not another warning but an AI-driven incident that hits a bank or critical infrastructure in a way markets can measure. Van der Chijs offered no specific technical examples, so any real-world disruption would be the first hard data point that turns his thesis from narrative into a risk premium.
On the corporate side, traders will be watching whether AI infrastructure continues to outcompete mining for capital at firms like Hut 8, which would reinforce his claim that AI data centers have the stronger business case. If that capital allocation trend persists, it can change how the market values crypto-adjacent infrastructure equities relative to pure-play mining exposure.
The other unresolved piece is whether the “AI capital diverted from BTC” story gains broader traction with measurable flow data. Van der Chijs’s price-cap range of $200,000 to $250,000 was presented as his interpretation, and it will stay that way until the allocation shift shows up in fund flows or in repeated disclosures from other large allocators.
My Read: This Is a Sentiment Signal, Not a Catalyst—Until the Flows or Incidents Show Up
The AI warning will get the headlines, but it is the portfolio tell that traders can actually use: he is explicitly rotating some AI profits back into bitcoin exposure via ETFs after saying he sold a lot of BTC to fund AI. That frames AI and BTC as competing buckets in a way that matches how many discretionary books already behave, even if they rarely say it out loud.
The threshold that matters is whether this turns into observable behavior beyond one investor’s rebalance, either through disclosed ETF choices and growing allocation size, or through flow and incident data that forces the systemic-risk narrative into pricing. Until one of those shows up, this reads more like a sentiment signal than a catalyst, and it only becomes practical if it changes where capital actually lands.
