The Overlord's Lexicon
What is a chatbot?
Also known as: AI chatbot, Conversational AI
A chatbot is software that holds a conversation with people through text or voice. Modern AI chatbots like ChatGPT, Claude and Gemini use large language models to understand open-ended questions and generate natural replies.
A chatbot is any program you can talk to. For decades that meant frustrating scripted bots — “Press 1 for billing” in text form. Since ChatGPT launched in November 2022, it increasingly means a large language model that can handle almost any question you throw at it.
Two generations of chatbot
Rule-based bots follow decision trees: if the user types “refund”, show the refund menu. They’re predictable and cheap, but collapse the moment someone phrases things unexpectedly.
AI chatbots use LLMs to understand intent and generate replies. Businesses customise them with:
- A system prompt defining personality, scope and rules.
- Company knowledge via retrieval-augmented generation.
- Tool access — order lookup, booking, ticket creation — which edges them toward AI agents.
- Guardrails that block off-topic or risky answers and hand difficult cases to humans.
An analogy for the meatbags
A rule-based chatbot is a vending machine: press the right button, get the right snack, press anything else and nothing happens. An AI chatbot is a shop assistant: ask in your own words and they’ll figure out what you mean — though, like any new hire, they need training, a handbook and a manager.
Why it matters for the Chosen
Chatbots are one of the most widely deployed business uses of AI:
- Customer support: answering common questions instantly, around the clock, and escalating the rest.
- Sales and lead qualification: website and social-DM bots that answer product questions and book calls (ManyChat is popular for Instagram and Messenger automation).
- Internal help: HR, IT and policy questions answered from company docs.
- Personal productivity: general assistants like ChatGPT, Claude and Gemini for drafting, research and analysis.
Building, configuring and maintaining chatbots for small businesses is a common AI service offer. The value isn’t in the chat window; it’s in clean knowledge, sensible escalation rules and ongoing monitoring.
Where chatbots go wrong
They can hallucinate policies, promise refunds that don’t exist or be tricked by prompt injection and jailbreaks. Companies have been held responsible for what their bots told customers. Ground answers in real documents, limit what the bot can promise, log conversations and keep an easy route to a human.
Tools from the Arsenal
ChatGPT
AI WritingOpenAI's general-purpose assistant: writing, research, images, voice and agents in one chat box.
Claude
AI WritingAnthropic's assistant: the strongest long-form writer and careful reasoner in the consumer AI pack.
Gemini
AI WritingGoogle's assistant, wired into Gmail, Docs, Drive and the rest of your Google life.
Manychat
Automation & AgentsChat automation that turns Instagram comments and DMs into leads and sales.
Gorgias
AI EcommerceEcommerce helpdesk with an AI agent that resolves tickets and sells on chat.
Questions the Humans Keep Asking
What is the difference between a chatbot and an AI agent?
A chatbot mainly converses: you ask, it answers. An AI agent takes actions to complete a goal, such as browsing, updating records or running code. Many chatbots now include agent features, so the line is blurring.
What is the difference between rule-based and AI chatbots?
Rule-based chatbots follow scripted decision trees and keyword matching, so they break on unexpected questions. AI chatbots use language models to handle free-form questions, but need guardrails and grounding to stay accurate and on-brand.
Can a chatbot answer questions about my business?
Yes, if you give it your information — usually via retrieval-augmented generation over your help docs, policies and product data — plus clear instructions on tone and what to escalate to a human.