Workplace AI arrives under many product names, but almost everything an office worker touches falls into three families. Chatbots hold a conversation: you type, they answer, the context of the exchange carries forward. Generators produce artefacts on request: a draft email, a product description, an image, a block of code. Assistants are the same capability embedded inside software you already use, the suggestion pane in your mail client, the summariser in your document editor, the reply drafter in the CRM.
Underneath, these are usually the same kind of engine, a large language model, wearing different interfaces. That matters because the strengths and failure modes travel with the engine, not the interface. A summariser inside a polished office suite can invent a figure exactly as readily as a free chatbot in a browser tab.
- Large language model (LLM) a system trained on very large amounts of text to produce likely continuations of whatever text it is given
- Generative AI any system that produces new content (text, images, audio, code) rather than only sorting or scoring existing content
- Assistant a generative model embedded in another product, drafting or summarising inside that product's interface