ARTICLE AD BOX
In a surprising revelation, OpenAI CEO Sam Altman shared that a single ChatGPT query uses a few drops of water. This comes at a time when the environmental cost of artificial intelligence is under growing scrutiny. In a blog post, Altman said each query consumes about 0.000085 gallons of water. That’s roughly one-fifteenth of a teaspoon. AI models like ChatGPT run on massive server farms that must be cooled constantly. This makes water usage an important part of the conversation. Altman’s claim aims to ease public concern, but some experts want more clarity and proof.
How water usage is connected to ChatGPT
AI runs on powerful computers stored in data centers that produce a lot of heat. To keep them from overheating, companies use cooling systems that often depend on water. As tech becomes more central to daily life, water use has joined energy and carbon emissions in the sustainability debate.
Sam Altman’s water estimate and what it means
Altman said each ChatGPT query takes about 0.34 watt-hours of electricity and a few drops of water. That may sound small, but when you think about the millions of queries made each day, the total adds up. Critics point out that OpenAI has not explained how this number was calculated. That lack of detail has made some experts cautious.
Past concerns about AI’s water use
A report from The Washington Post last year estimated that creating a 100-word email with GPT-4 could use more than a full bottle of water. These numbers were tied to the cooling needs of data centers, especially those in hot and dry places. Altman’s latest statement appears to push back on that report as pressure grows on tech firms to be more accountable.
Experts call for transparency
Many in the tech and environmental space say companies like OpenAI need to publish independent and verified data about their resource use. Altman’s number sounds reassuring but without knowing how the math was done or where the servers are located, it is hard to trust fully.
Can AI be sustainable?
As AI becomes a part of more industries and daily life, its long-term environmental cost matters more than ever. Altman believes the cost of intelligence will one day drop to the price of electricity alone. That could make AI both affordable and sustainable. But for now, even few drops of water per query raise big questions.