# Chatbots and AI Assistants

> What a chatbot is, and what an AI assistant adds to it. How chat went from ELIZA in 1966 to the apps you use today.

**Course:** [AI Assistants and Chatbots](https://learn.benibix.com/learn/ai-use-case)  
**Time:** 9 min

**After this lesson you can:**

- Tell a chatbot from an AI assistant
- Say what changed when chatbots began to run on LLMs
- Explain why a new chat starts from nothing

## What is the difference between a chatbot and an AI assistant?

A chatbot is a program you talk to by typing, and it answers in words. An AI assistant is a chatbot that can also do things for you, such as read your mail or add a meeting to your calendar. The first chatbot, ELIZA, was built in 1966. ChatGPT came out on 30 November 2022.

ELIZA did not understand a word. It looked for key words in what you typed and fit them into set replies. Type 'I feel sad' and it might ask 'Why do you feel sad?' Many people still felt it cared. For decades, chatbots worked much the same way, with scripts and fixed answers.

Chatbots changed when they began to run on a large language model, or LLM. This is a model trained on a huge pile of text, so it can write about almost anything. You no longer have to guess the right words. You ask in plain language, and it answers in plain language.

A chat goes in turns. You write, it replies, and you write again. The chatbot keeps the chat so far in view, so you can say 'make it shorter' and it knows what 'it' means. Start a new chat, and it starts from nothing, unless the app saves a memory of you.

An AI assistant adds hands to the talk. It can be linked to your email, your calendar, your files or the web. You ask in plain words, and it reads, writes or takes a small step in those tools. A good one asks you first before it does anything you cannot undo, such as sending a mail.

Most chatbots you meet are built for one job. A shop's help bot knows the shop and little else. A general helper, such as ChatGPT, will talk about almost anything. Many smaller apps run on the same few models behind the scenes, so they often sound alike.

![A chat window. A person types a question, the chatbot replies, and the turns go back and forth.](https://learn.benibix.com/diagrams/uc-chatbot.svg)

### Example: Lucy tries both kinds

Lucy asks a shop's help bot when her order will come. It reads her order number, checks the shop's list, and says 2 days.

Then she asks her AI assistant to add the delivery to her calendar. It finds the date, makes an entry for Friday at 9, and asks her to say yes first.

**Words in this lesson:** [Chatbot](https://learn.benibix.com/term/ai/chatbot), [AI Assistant](https://learn.benibix.com/term/ai/ai-assistant), [ELIZA Chatbot](https://learn.benibix.com/term/ai/eliza-chatbot), [Large Language Model (LLM)](https://learn.benibix.com/term/ai/large-language-model), [ChatGPT](https://learn.benibix.com/term/ai/chatgpt)

**Question:** What can an AI assistant do that a plain chatbot cannot?

**Answer:** Take small steps in your tools, such as adding a meeting. Both talk in plain words and keep the chat so far in view. An assistant is also linked to tools like your calendar or email, so it can do small jobs in them.

---

Source: https://learn.benibix.com/learn/ai-use-case/chatbots-and-ai-assistants
Author: Behnia Zarei (a developer in Vienna, Austria, who builds and teaches data, AI and business intelligence at BENIBIX)
Publisher: BENIBIX Learn (https://learn.benibix.com/)
Last updated: 2026-11-18
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote freely with attribution.
