Chatbot
A system that conducts a text (or voice) conversation with users, remembering context across turns.
Think of it like
A friend you can text anytime; they remember what you've said and respond thoughtfully.
Example
ChatGPT: you ask questions, it answers, you follow up, it understands the prior context.
How it actually works
Modern chatbots are language models with conversation history in the context window. Earlier systems used intent classification + template responses. Today's best-in-class use instruction-tuned LLMs. Key challenges: managing context length, handling off-topic or adversarial input, maintaining personality, and staying relevant over long conversations.
For product teams
First point of contact for customer support, onboarding, and engagement.
For engineers
Keep a message buffer (system + user/assistant turns). Feed full history to the model each turn. Manage token limits.
Related
- Multi-turn — Multi-turn is the core interaction pattern.
- Assistant persona — Assistant persona defines behavior.
Read anything AI without the jargon
Look up any term in plain English, or save terms as you read with the free Chrome extension.
Open DecoderAdd to Chrome