When AI CEOs Compare Chatbots to Human Children

At a major AI summit in India last Friday, a reporter from The Indian Express posed a pointed question to Sam Altman: What about the natural resources required to train and run generative AI models?

The OpenAI CEO didn't hesitate. "It also takes a lot of energy to train a human," he told the packed pavilion. "It takes like 20 years of life and all of the food you eat during that time before you get smart." He went further, invoking the "very widespread evolution of the hundred billion people that have ever lived" before concluding that AI has "probably already caught up on an energy-efficiency basis" compared to humans answering questions.

The Flawed Math Behind the Comparison

Altman's analogy sounds clever, but it doesn't hold up to scrutiny. The human brain uses significantly less energy than even the most efficient AI models for simple queries—not to mention the laptops and smartphones people need to access these chatbots in the first place.

More importantly, Altman's deflection misses the real concern. The issue isn't energy consumption per se, but climate impact. Atmospheric carbon dioxide levels haven't been this high in millions of years—driven not by human evolution, but by contemporary society's combustion turbines, like those OpenAI is installing at its Stargate data centers.

Other tech companies are building private gas-fired power plants that will collectively generate enough electricity—and emit as much greenhouse gas—as dozens of major American cities. Some are even extending the life of coal plants.

When Machines Become People

What's truly significant about Altman's remarks isn't their factual accuracy, but that he thought to compare chatbots to humans at all. This suggests he views people and machines on fundamentally equal terms—and this isn't a slip of the tongue.

Altman made nearly identical comments to Forbes India at the same summit. A week earlier, Dario Amodei, CEO of Anthropic and Altman's chief rival, drew similar parallels between AI training and human evolution. The mindset trickles down to product development: Anthropic studies whether its chatbot Claude can feel "distress" and allows it to cut off conversations when there are "risks to model welfare."

This anthropomorphization of software that doesn't eat, drink, or have any will of its own reveals a troubling worldview.

God Complex or Marketing Genius?

AI firms seem convinced either that their products really are comparable to humans, or that this makes for compelling marketing. Both possibilities are alarming.

If tech leaders genuinely believe they're building a higher power—Altman said superintelligence is just years away—they might easily justify treating humans and the planet as collateral damage. When pressed about energy consumption, Altman argued that since "the world is now using so much AI," societies must "move towards nuclear, or wind and solar, very quickly." Notably absent: the option for the AI industry to simply slow down.

If it's pure PR, the strategy is deeply misanthropic. Altman was speaking to investors, after all. The notion that AI labs are creating digital life has always been convenient to their mythology—and OpenAI is reportedly raising funds at an $800 billion valuation, nearly matching Walmart.

What We Lose When Everything Becomes Efficient

Tech companies may genuinely want to develop AI tools for humanity's benefit, echoing OpenAI's founding mission. They may sincerely believe they need massive capital to do so. But likening raising a child—or human evolution itself—to developing algorithmic products reveals how disconnected the industry has become from what it means to be human.

To "train a human"—to live a life—is to struggle, accept the possibility of failure, and sometimes wander in search of wonder and beauty. Generative AI aims to eliminate that process, making every pursuit as instant, efficient, and effortless as possible.

These tools may serve us well. But placing them on the same plane as organic life diminishes something essential about the human experience.