Sam Altman would like remind you that humans use a lot of energy, too

OpenAI CEO Sam Altman has launched a forceful rebuttal against growing concerns about the environmental footprint of artificial intelligence, specifically targeting viral claims about ChatGPT's resource consumption. Speaking at an event hosted by The Indian Express during a major AI summit in India, Altman dismissed widespread internet assertions that 'Don't use ChatGPT, it's 17 gallons of water for each query' as 'completely untrue, totally insane, no connection to reality.' He clarified that while early data centers used evaporative cooling—a water-intensive process—modern systems have moved beyond this method, making current water usage claims inaccurate.

**Background/Context:** The debate about AI's environmental cost has intensified as models like GPT-4, Claude 3, and Llama 3 have scaled to consume massive computational resources during training. With no legal requirements for tech companies to disclose energy and water usage, independent researchers have attempted to quantify the impact. Studies have suggested data centers contribute to rising electricity prices, and comparisons have circulated claiming a single ChatGPT query uses energy equivalent to 1.5 iPhone battery charges—a claim Altman directly refuted, stating 'There's no way it's anything close to that much.'

**Technical Details:** Altman made a crucial distinction between training energy (the one-time cost to create a model like GPT-4) and inference energy (the cost per query). He argued that focusing solely on training costs creates an 'unfair' comparison. Instead, he proposed evaluating efficiency by comparing the energy for a trained AI to answer a query versus a human performing the same task. From this perspective, he suggested 'AI has already caught up on an energy efficiency basis.' He emphasized that the real concern should be 'the energy consumption—not per query, but in total, because the world is now using so much AI.'

**Impact Analysis:** Altman's comments represent a strategic shift in how AI companies address sustainability concerns. By challenging specific viral claims and reframing the efficiency debate, OpenAI aims to deflect criticism while acknowledging the broader issue. His advocacy for 'nuclear or wind and solar' energy solutions aligns with tech industry trends toward power purchase agreements for renewables. However, without transparent data from OpenAI about actual energy consumption per query or total data center usage, independent verification remains impossible. The discussion highlights the tension between rapid AI adoption and responsible environmental stewardship.

**Future Implications:** Altman's framing suggests AI companies will increasingly argue that AI represents a net efficiency gain for society despite high upfront training costs. This could influence policy debates about data center regulations and carbon accounting. His call for accelerated clean energy adoption indicates that major AI players recognize their growth depends on abundant, sustainable power. As models grow larger—with rumors of GPT-5 requiring even more resources—the industry will face mounting pressure to either prove its efficiency claims or invest substantially in offsetting its environmental impact. The conversation also sets the stage for potential standards around AI energy disclosure, similar to carbon reporting requirements in other industries.