Sam Altman: AI Under Scrutiny for Profitability

Sam Altman and the Profitability of AI Investments

Sam Altman, CEO of OpenAI, recently addressed what he considers to be "the most valid criticism" of artificial intelligence: the return on investment from massive spending in this field. In an interview on CNBC, Altman acknowledged that investors' concerns about the profitability of AI investments are legitimate, as companies continue to spend billions on infrastructure, chips, and software.

"I think it's the most valid criticism of AI right now," Altman stated. He emphasized that while significant advancements are being made, there is also considerable waste. Companies are questioning how long it will take before these investments translate into concrete revenue and how to manage costs.

Profitability Challenges for OpenAI

Altman's comments come at a time when investors are increasingly scrutinizing whether the AI boom can generate enough revenue to justify its exorbitant costs. According to an April report from the Wall Street Journal, OpenAI itself missed several key revenue and user growth targets last year, adding pressure on the company to prove the economic viability of its projects.

Inefficient Use of AI Resources

Altman's concerns are reinforced by recent data from Cast AI, a cloud optimization platform. According to their analysis of 23,000 clusters across thousands of companies, the average GPU utilization is only 5%, leaving about 95% of the provisioned graphics processing capacity unoccupied. Laurent Gil, co-founder of Cast AI, explains that companies often accumulate rare AI chips out of fear of missing an opportunity, rather than immediate need.

Colossal Investments Without Guaranteed Returns

Gary Marcus, an AI researcher and emeritus professor at New York University, has also criticized the massive investments in AI. He described these expenditures as "the largest misallocation of capital in history." Marcus pointed out that companies like Amazon, Google, Microsoft, and Meta collectively spend more each month on AI than what was spent on the Manhattan Project.

These observations highlight the need for the AI industry to demonstrate tangible returns on investment to justify the exorbitant costs incurred.