Oracle hit hard in Wall Street’s tech sell-off over its huge AI bet (arstechnica.com)

Oracle’s aggressive pivot into AI — committing to spend “hundreds of billions” on chips and data centers to supply compute capacity to OpenAI — has unnerved investors and driven a sharper sell-off than its hyperscaler peers. Shares fell about 25% in the past month (nearly double the next worst performer) and have wiped out roughly $250 billion in market value since the OpenAI deals were disclosed; an FT index tracking Oracle’s debt is down ~6% since mid‑September. The market reaction reflects worry that Oracle’s move is financed by heavy borrowing and is a departure from the lower‑capex, subscription‑driven cloud model investors prefer. For the AI/ML community this is significant because it signals a major reconfiguration of large‑scale compute supply: Oracle is betting on building massive GPU/accelerator farms and data‑center capacity to host training and inference workloads (notably for OpenAI). That creates potential benefits — more competition for hyperscale compute, possibly more specialized capacity — but also risks: the model is extremely capital‑intensive and tightly coupled to the commercial success of loss‑making AI startups. If customers like OpenAI or Anthropic don’t monetize advanced models as hoped, Oracle’s returns could underperform, affecting pricing, availability of training capacity, and the broader economics of large‑scale model development.