ChatGPT

When OpenAI released ChatGPT in November 2022, it did something few AI products had managed at that scale — it made the technology feel immediately useful to ordinary people. You did not need a technical background to get value from it. You typed something in plain English, and something genuinely helpful came back. Within weeks of launch it had attracted a user base that dwarfed anything the AI industry had seen before, becoming among the fastest-adopted consumer products in history. For many people, ChatGPT was their first real encounter with what modern AI could do, and it permanently shifted expectations about what software was capable of.

Under the hood, ChatGPT runs on a large language model — the GPT in the name stands for Generative Pre-trained Transformer, a reference to the type of AI architecture it is built on. You do not need to understand what that means to use it effectively, but if you are curious, the Transformer entry in this glossary explains it. What made ChatGPT different from earlier AI chatbots was not just the underlying model but how it had been trained to behave. OpenAI used a technique called reinforcement learning from human feedback, where human trainers rated the model's responses and those ratings were used to steer it toward answers that were more helpful, more consistent, and less likely to cause harm. The result was a system that felt less like querying a database and more like talking to a knowledgeable colleague.

Since its launch, ChatGPT has grown into a platform in its own right, expanding well beyond text conversations to handle images, files, and external data, with new capabilities added regularly. Competing AI assistants from Google, Anthropic, Meta, and others have followed, but ChatGPT remains the most recognized name in AI assistants for most general users. The product that introduced most people to the generative AI era — whether you use it for drafting emails, debugging code, researching a topic, or just thinking through a problem out loud.