LLM chatbot

Chatbots had existed for decades—ELIZA in 1966, customer service bots in the 2000s, Siri and Alexa in the 2010s. But these systems followed scripts or matched patterns; they couldn't hold genuine conversations or answer unexpected questions. Large language models had demonstrated remarkable capabilities since GPT-3 in 2020, but they remained tools for developers, accessible through APIs. The missing piece was a user interface that would let anyone experience what these models could do.

On November 30, 2022, OpenAI released ChatGPT. It was, at its core, a simple product: a text box where users could type anything and receive conversational responses from an LLM fine-tuned for dialogue. The simplicity was the innovation. No API keys, no programming knowledge, no subscription required. Ask it to explain quantum physics, write a poem, debug code, or role-play as a pirate—the model would attempt it all. Within five days, ChatGPT had reached 1 million users. Within two months, 100 million. No technology product in history had grown so fast.

The adjacent possible was precisely configured. GPT-3.5 provided the underlying capability. Reinforcement learning from human feedback (RLHF) had made the model's outputs more helpful and less harmful. Cloud infrastructure could scale to millions of simultaneous users. The technology was ready, but the key insight was democratizing access through a simple chat interface rather than a developer tool.

The cascade was immediate and industry-reshaping. Google rushed Bard to market. Anthropic launched Claude. Microsoft invested $10 billion in OpenAI and integrated GPT into Bing and Office. Meta released open-weight models. Every major tech company scrambled to incorporate 'AI assistants' into their products. Schools debated how to handle AI-written essays. Professionals wondered which jobs would transform or disappear. Governments struggled to regulate a technology that was evolving faster than policy.

By 2025, LLM chatbots had become a new computing interface—joining command lines, graphical interfaces, and touchscreens as ways humans interact with machines. The technology that seemed exotic in late 2022 had become infrastructure: embedded in search, productivity tools, customer service, and education. ChatGPT's simple innovation—making LLM capabilities accessible to everyone—had triggered one of technology's most rapid transformations.