Sam Altman says a whole generation of researchers held AI back by underestimating what scaling could do

At a recent talk at Stanford, Sam Altman emphasized the importance of scaling in AI, arguing that many researchers have historically underestimated its potential, which he claims has hindered progress in the field. He referenced OpenAI's successful disproof of a mathematical conjecture as a demonstration of the capabilities unlocked by larger models. This perspective highlights the significance of scaling laws for practitioners working with LLMs, suggesting that further advancements may rely on continued investment in model size and complexity.