Sam Altman says OpenAI isn't trying to become 'too big to fail' (www.businessinsider.com)

OpenAI CEO Sam Altman pushed back on suggestions the company is seeking government guarantees after CFO Sarah Friar walked back comments about a “government backstop.” In posts on X he said OpenAI “does not have or want” taxpayer bailouts, warned governments shouldn’t pick winners, and repeatedly insisted the company is not trying to become “too big to fail.” Altman defended the company’s aggressive infrastructure build‑out — citing recent multi‑billion compute deals and an announced plan to spend as much as $1.4 trillion over the next eight years — by pointing to urgent demand and long lead times for data centers. He said OpenAI expects a $20 billion annualized revenue run rate by year‑end and sees enterprise services, consumer devices and robotics driving growth that could scale into the “hundreds of billions” by 2030. For the AI/ML community, the remarks crystallize two tensions: the raw economics of training and serving large models, and the political/regulatory scrutiny that follows extreme concentration of compute and talent. Technically, OpenAI’s posture signals continued massive investment in bespoke data‑center capacity and long‑horizon compute commitments, which accelerates the infrastructure arms race (and potential vendor lock‑in) for large models and AGI research. Altman’s note that AI is a “national strategic asset” also signals governments may invest directly in their own compute, raising questions about regulation, competition, and who controls critical AI infrastructure as models scale.