Sam Altman: AI Soon to Be as Essential as Water

⚡

Key Takeaways

1Sam Altman, CEO of OpenAI, predicts that AI will be sold as a public utility, measured by usage.

2Computing power is crucial for access to AI, and demand continues to grow, according to Altman.

3Tech companies are investing heavily to increase their computing capacity to meet this demand.

💡Why it matters — The distribution of AI as a public utility could transform access to technology, influencing global infrastructure and energy policies.

AI: A New Utility Bill?

Sam Altman, the CEO of OpenAI, recently shared a bold vision for the future of artificial intelligence (AI). He envisions a world where AI is sold as a public utility, similar to electricity or water. This approach would transform AI into a commodity measured by usage, making access as commonplace as other essential services.

Altman emphasized that the demand for AI is rapidly expanding. According to him, computing capacity—the power needed to train and run AI models—will be the determining factor for who can utilize this technology. This capacity is particularly influenced by infrastructures such as chips and data centers.

A Shared Vision at the BlackRock Summit

At the BlackRock Infrastructure Summit in Washington, DC, Altman explained that companies like OpenAI are working towards a future where intelligence is provided on demand. He compared the sale of AI to that of "tokens," the units that AI systems use to process data. Altman clarified that this approach could lead to a situation where computing capacity becomes a critical issue.

If OpenAI fails to develop sufficient capacity to meet the growing demand, Altman fears that the company may be unable to sell its services or that prices could become prohibitive. This could limit access to AI to only the wealthiest entities or force governments to intervene to regulate the distribution of this limited resource.

Massive Investments in Infrastructure

Major tech companies are already spending hundreds of billions of dollars to increase their computing capacity. At CES 2026, Lisa Su, CEO of AMD, estimated that the world will need over 10 yottaflops of computing capacity in the next five years, a massive increase compared to the global capacity in 2022, which would be 10,000 times greater.

This rapid growth poses significant challenges in terms of infrastructure. The data centers required to support AI consume enormous amounts of electricity, sometimes as much as small cities. In the United States, this could create strains on the electrical grid, which is already facing shortages of transformers and delays in the approval of new transmission lines.

The Race for Energy and Computing Capacity

Elon Musk, in a recent podcast, highlighted that electricity production is now the main limiting factor for the expansion of AI. He predicted that China could surpass the United States in total AI capacity, thanks to its rapid energy development.

In the tech sector, computing capacity has become a valuable resource. Engineers are competing for access to GPUs, and some job candidates are now asking for details about the AI computing budget in addition to their compensation.

OpenAI and Its Ambitions for the Future

Greg Brockman, president of OpenAI, revealed that the company has committed approximately $1.4 trillion to data center projects over the next eight years. Despite these colossal investments, Brockman remains cautious about OpenAI's ability to stay ahead of the demand curve.

At the BlackRock Summit, Altman expressed his desire to move away from a world where access to AI is "constrained by capacity." This ambition reflects a willingness to democratize access to AI while highlighting the logistical and economic challenges that this entails.