The AI Bubble Might Not Exist After All
The author challenges the prevailing assumption that the AI industry is in a bubble. While acknowledging the astronomical spending—estimated at 5 trillion dollars total—he argues this reflects justified capital investment in infrastructure rather than unsustainable operating expenses. Once data centers are built, the ongoing costs to run AI systems are manageable; Anthropic’s filings show the company generates $2.08 in revenue for every $1 spent on compute, and has started exceeding combined infrastructure and compute costs. Even debt repayment of roughly $250 billion annually over 20 years is theoretically manageable given current revenue trajectories.
The case for continued growth rests on massive unmet demand. AI companies have demonstrated explosive revenue growth—Anthropic grew from $41 million in 2023 to $16.3 billion in just the first half of 2026. Applications remain nascent across programming, medicine, legal work, and autonomous vehicles. The author notes skeptics contradict themselves by claiming AI will transform society while simultaneously dismissing it as worthless—money inevitably flows toward capable tools. Though companies face no durable competitive advantage, the author argues this doesn’t prevent profitability, drawing a parallel to fast-food chains during the chicken sandwich wars. Users tend to stick with established products despite theoretical ability to switch, creating practical competitive advantages that support profitable scale.
Read the full article
Continue reading at Hacker News →This is an AI-generated summary. Read the original for the full story.