OpenAI released ChatGPT as a free research preview, a conversational interface built on a GPT-3.5-series model. It could answer questions, write essays and code, and hold extended back-and-forth conversations — and it exploded, reaching an estimated 100 million users within two months, the fastest-growing consumer app in history at the time.

The launch kicked off the modern AI race overnight. Google declared a 'code red' within weeks, Microsoft deepened its OpenAI partnership, and every major lab reorganized around conversational AI.

ChatGPT's underlying model was GPT-3.5, an instruction-tuned evolution of GPT-3. It set the template for everything that followed: the chat interface, RLHF-tuned assistants, and the Plus subscription tier that arrived the following February.