By Billy Perrigo
Publication Date: 2025-12-11 12:40:00
OpenAI, which sparked the boom, continues to set the pace in many ways. ChatGPT usage has more than doubled, to 10% of the world’s population. “That leaves at least 90% left,” says Nick Turley, head of ChatGPT.
Read more: Why the architects of AI are TIME’s Person of the Year 2025
A large language model (LLM), the technology underlying chatbots like ChatGPT or Anthropic’s Claude, is a type of neural network, a computer program that is different from typical software. By feeding massive amounts of data, engineers train the models to recognize patterns and predict which “tokens,” or word fragments, should come next in a particular order. From there, AI companies use reinforcement learning—strengthening the neural pathways that lead to desired responses—to turn a simple word predictor into something like a digital assistant with a finely tuned personality.
About a year ago, OpenAI researchers found a new way to improve these models. Instead of letting them respond to requests…
