Sam Altman Addresses Investor Concerns Over AI Spending

According to Business Insider's report of a CNBC interview, OpenAI CEO Sam Altman said questions about whether AI spending will pay off are "the most fair criticism right now." Business Insider quotes Altman: "I am spending a ton of money on AI. And I know some great stuff is happening, but I know there's a ton of waste." He added: "How long do I have to wait for it to really show up in revenue, and how long do I have to wait to really get the costs under control?"

Business Insider also cites an April report from The Wall Street Journal that OpenAI missed some key targets for revenue and user growth last year. Business Insider reports that an analysis covering 23,000 GPU clusters across thousands of companies found average GPU utilization of just 5%, meaning roughly 95% of provisioned GPU capacity was idle in that sample.

Industry-pattern observations: large-scale AI deployments drive up capital and operating expenditures because training and serving models require dense GPU capacity, specialized networking, and storage. Observers tracking cloud and on-prem GPU fleets have repeatedly reported low utilization in early AI rollouts as teams overprovision for peak load and experimentation.

Altman's acknowledgment matters because it echoes investor scrutiny over AI economics, particularly unit economics of running large models and the timeline for revenue realization. Reporting that OpenAI missed some targets, and that GPU utilization can be low, underscores why investors and customers are asking when AI investments translate into consistent revenue and higher infrastructure efficiency.

For practitioners: monitor published utilization telemetry and cost-per-inference metrics from cloud providers and optimization vendors. Industry observers will also track vendor earnings and public reports for signs of improving GPU utilization, decreasing cost-per-query, or new pricing/ops patterns that shift the cost calculus for AI projects.