Decode the Jargon of Artificial Intelligence
Decode the Jargon of Artificial Intelligence
A Curated Glossary of Common Artificial Intelligence Jargon, for Project Managers
Now in it Edition 5.2 (12 Jan 2026), with a total of 167 definitions and an enlarged (52 entries) Acronym Decoder.
The fifth edition of this glossary was a substantial increment (over 20%) on the previous version, 4.5. It contained an additional 30 terms and 13 acronyms that were not in the previous edition.
Editions since 4.0 have been AI-enabled. Once again, for the fifth edition (5.0), I asked the latest version of ChatGPT (5), for new jargon, terminology, and acronyms that have appeared since the start of 2025. It suggested 6 terms I had not collected myself.
Edition 5.2
With only an additional 6 definitions and 2 more acronyms, the reason I put version 5.1 out so soon after 5.0 was the term AI Psychosis. It’s not a clinical term, but it is strongly indicative of the way AI is affecting society.
This edition, number 5.2, of Decode the Jargon of Artificial Intelligence adds 5 more terms, including AI Slop and Context Engineering. It also adds a new Special Feature, about Small Language Models (SLMs). This version contains 167 definitions and 53 acronyms. Please get in touch if there are any terms or acronyms you think I should add to this glossary.
This version of Decode the Jargon of Artificial Intelligence now includes 6 Special Features and explainers interspersed among the glossary items:
- Understanding Differences: AI Agents and RPA
- Understanding Benchmarks
- Forms, Causes, and Fixes of Bias in Artificial Intelligence
- The Layers of a Large Language Model (LLM)
- Types of Machine Learning
- SLM vs purpose build GPT
Please get in touch if there are any terms or acronyms you think I should add to this glossary.