AI has quickly become part of how businesses talk to their customers. At the same time, the terms get mixed up: artificial intelligence, AI chat, ChatGPT and chatbots are often used for the same thing, while they mean something different.
This guide sets out the terms and shows what they mean for customer contact. For each topic we point you to an article that goes into more depth.
Artificial intelligence: the basis
Artificial intelligence, or AI, is technology that carries out tasks that normally require human thinking: understanding language, recognising patterns and preparing decisions.
Most of the AI you come across in customer contact is based on machine learning. A system learns from large numbers of examples rather than from fixed rules. Language models are one form of this: they are trained on text and can therefore understand questions and formulate answers. How AI works and where businesses use it is covered in what is artificial intelligence.
AI chat: conversations with a language model
AI chat is a conversation between a person and an AI system in plain language. You type a question and the system formulates an answer, instead of sending you a list of search results.
In customer contact this usually happens in a chat window on the website or in WhatsApp. The difference from a general AI assistant is the information the answer is based on: in customer contact, the answer has to match your products, prices and terms. What AI chat is and how to use it is covered in what is AI chat.
ChatGPT and similar assistants
ChatGPT is an AI assistant from OpenAI that runs on a language model. It helps with writing, summarising, explaining and brainstorming, and for many people it is their first encounter with AI chat.
ChatGPT knows a lot in general, but nothing about your organisation unless you give it that information yourself. It can also sound convincing when an answer is wrong. What ChatGPT can do and where its limits lie is covered in what is ChatGPT.
ChatGPT in your organisation
Many teams use ChatGPT internally: for texts, summaries of documents or preparing answers. That can save a lot of time, provided you agree what goes into it and who checks the result.
For customer conversations, a standalone AI assistant is less suitable. Customers expect an answer that matches your offering, and a person when the question gets too complicated. For that you need a solution that is set up on your information, connected to your systems and monitored in practice. The options, limits and alternatives are covered in ChatGPT for business.
From AI chat to chatbot and AI agent
AI chat is the technology; a chatbot is the application built with it for customer contact. A chatbot answers questions in a fixed channel, with information about your organisation.
An AI agent goes a step further. It does not just answer questions, but also carries out tasks, such as booking an appointment, looking up a status or passing a lead to the right team. More about chatbots is in the guide to chatbots.
Where AI works in customer conversations, and where it does not
In the conversations we handle at Bconnect, we see that AI is strong at predictable questions: opening hours, delivery times, the status of an appointment or a question that is already in the knowledge base.
With advice, bespoke work and valuable sales opportunities, it is different. There a chat specialist who asks follow-up questions makes the difference. That is why we never let AI be the end point: if the AI cannot help, it announces the handover and the conversation goes to a person with the full context.
That way you can use AI in your customer contact without building it yourself: we set up the AI on your information, connect it to your systems and keep managing it.