Matt Wolfe

Sam Altman Compares Training AI To Raising Kids

In a viral clip, OpenAI CEO Sam Altman likens the process of training large language models to raising a child, framing AI development as a form of digital parenting. Altman points out that while critics focus on the electricity required to train a model, humans also expend massive energy over decades of growth, nutrition, and education. He argues that once a model is trained, the per‑query energy cost may already match or surpass human efficiency. He illustrates his point with a vivid analogy: “It takes twenty years of life and all the food you eat before you get smart,” and reminds viewers that humanity’s collective evolution—hundreds of billions of ancestors—underlies our intelligence, a cost AI sidesteps. If AI indeed achieves comparable energy efficiency, businesses could see lower operational expenses and faster scaling, while policymakers must grapple with the environmental and ethical implications of accelerating AI deployment.