Sam Altman’s vision for AI is huge – but there’s just one thing standing in his way (www.techradar.com)

OpenAI CEO Sam Altman’s new blog post, "Abundance Intelligence," lays out a stark, infrastructure-first roadmap: build a "factory" that can add a gigawatt of AI infrastructure every week. By "compute" Altman means the warehouse-scale data center horsepower used to train and run large language models, and he’s already showing progress — a video from the Abilene, Texas site, part of the massive Stargate project, and a reported $100 billion Nvidia investment that will deploy power comparable to ten nuclear reactors underscore the scale. Altman frames this push as necessary to avoid rationing compute between competing uses (from curing disease to universal personalized tutoring) and pledges heavy domestic investment in chips, energy, robotics and facilities. The significance for AI/ML is twofold: technically, sustained exponential increases in available compute would enable much larger models and more ambitious AI workflows, but practically it forces trade-offs across energy, manufacturing and geopolitics. Altman admits execution is "extremely difficult" and will demand innovation across the stack, while critics note environmental impact and that more compute hasn’t yet delivered AGI. He also elevates access to AI as an economic driver—and possibly a future human right—signaling OpenAI’s intent to shape both the technical and policy landscape. Expect more concrete partner announcements and technical plans in the coming months.