Venture capitalist and early Bitcoin investor Chamath Palihapitiya ignited a fierce debate on Sunday by outlining what he sees as two major structural challenges for Bitcoin and the broader crypto market. In a post on X, the Social Capital CEO argued that the next marginal dollar of speculative capital is flowing into prediction markets and equities rather than cryptocurrencies. His second point claimed that the energy currently used for Bitcoin mining could generate “10–20x” more value if reallocated to powering AI infrastructure and tokens. “These changes feel structural, but I could be wrong,” Palihapitiya wrote.
The comments drew swift pushback from prominent Bitcoin advocates. Coinbase CEO Brian Armstrong called the observations interesting but said the liquidity shift is likely temporary. He acknowledged AI’s growing demand for computing resources may be more durable, yet stressed that “hash power or energy going to Bitcoin mining doesn’t determine its price (the network difficulty adjusts if miners go offline to keep the same pace of block mining).” Armstrong added that in the long term Bitcoin’s price reflects inflation fears, which show no signs of abating.
Strike CEO Jack Mallers took a harder line, asserting that money rotating into prediction markets, memecoins, or AI was never loyal Bitcoin demand and losing it changes little. Bitwise strategist Jeff Park dismissed the post, noting Palihapitiya was “about nine months late” in recognizing the trend. Macro analyst André Dragosch questioned whether the AI boom itself could prove temporary, potentially negating the capital-diversion concern.
Other voices offered nuanced takes. Bitwise’s Matt Hougan conceded the first point is true and helps explain reduced Bitcoin volatility, while the second problem is largely self-correcting. David Hernandez framed both issues as cyclical rather than structural, and James Van Straten pointed out that miners have been pivoting toward AI for roughly two years.
The debate is backed by concrete data. A CoinShares study cited in the discussion found the average cost to mine one Bitcoin for public miners was about $79,995 in Q4 2025, when BTC traded near $68,000–$70,000, resulting in a loss of roughly $19,000 per coin. In response, the mining industry has signed over $70 billion in AI and high-performance computing contracts. Meanwhile, prediction markets saw a record $28.4 billion in monthly trading volume in May, with Kalshi handling $17.3 billion and Polymarket $8.4 billion, marking four consecutive months of rising volumes—a clear sign that speculative liquidity has found an active alternative destination.