foundations · Level 2
AGI explained: what it means and what it does not
There is no single definition. Here are the main ones, how to read the headlines, and what matters in practice.
Start with the essentials
The short answer
AGI, or artificial general intelligence, usually means AI that can do most of the cognitive tasks a person can. There is no single agreed definition, so claims that AGI has arrived, or is coming by a set date, are contested and often opinion. What matters in practice is measurable capability, reliability, safety and governance.
What you will learn
- You will be able to explain why AGI has no single agreed definition, and describe the main definitions fairly.
- You will understand the Levels of AGI framework, including its separate scales for performance, generality and autonomy.
- You will be able to tell observed results, interpretations, forecasts and definitions apart in a headline.
- You will know what related terms such as ASI, alignment and autonomy mean, and how the EU AI Act relates to the debate.
- You will have written a working definition of your own, with a way to test it.
Who it is for
Anyone who keeps meeting the term AGI in news, conversation or work and wants a calm, accurate way to think about it. No technical background is needed.
Before you start
- None beyond curiosity. What is AI is helpful background.
Keep learning
The complete workbook
This workbook explains AGI without hype or doom. You will meet the main definitions, a 2023 levels framework from Google DeepMind researchers, the difference between narrow and general AI, and a method for reading headlines. It ends with regulation and your own working definition.
- 01A term with no single definitionIn the workbook · 1 exercise
Everyone uses the letters AGI. Not everyone means the same thing by them.
- 02Levels of AGI: one published frameworkIn the workbook · 1 exercise
Instead of a yes or no answer, one published paper proposes levels, so that people can say more precisely where a system stands.
- 03Narrow versus general, and why claims are contestedIn the workbook · Reading
Most AI is narrow. The interesting cases are the ones that are broad but uneven.
- 04Reading AGI headlines and timelinesIn the workbook · 1 exercise
Headlines about AGI mix four different kinds of statement. Learn to separate them and most of the confusion clears.
- 05Related terms, and what matters in practiceIn the workbook · Reading
A few neighbouring words are constantly confused with AGI. Then we turn to what matters whatever label is used.
- 06Your own working definitionIn the workbook · 1 exercise
A working definition is a tool for thinking. It lets you say precisely what you mean, and test it.
Also inside: a 9-point checklist, a glossary of 12 terms and 10 questions and answers to test yourself. 4 hands-on exercises, each with a worked answer at the back where the workbook gives one.
No login, no card, no account. Before the download we ask you to follow Mickai (two quick links). Free to download and use for personal learning, study groups and inside your own team. Please do not resell the workbooks or republish them as your own. Link people to trust-agent.ai instead.
Test yourself
Questions and answers
What is AGI?
AGI, or artificial general intelligence, usually means AI that can do most of the cognitive tasks a person can. There is no single agreed definition, so it is important to ask which meaning someone is using.
Has AGI been achieved?
It depends on the definition, and there is no agreed test. Some people argue current systems already qualify in part, others disagree. Treat any flat yes or no with caution and ask what was measured.
What are the Levels of AGI?
A 2023 paper by Google DeepMind researchers rates AI on performance (from Emerging to Superhuman) and generality (narrow or general), and separately on autonomy. It is a shared vocabulary, not a prediction of when any level will arrive.
What is the difference between narrow and general AI?
Narrow AI is built for one task or a small set, and may be brilliant at it. General AI handles a wide range of tasks, including unfamiliar ones. Today's chatbots are broad but uneven.
Is AGI the same as superintelligence?
No. AGI is broadly human-level capability. Artificial superintelligence, or ASI, would outperform all humans across a wide range of tasks. In the Levels framework, ASI is the top level.
Should I trust predictions about when AGI will arrive?
Treat them as opinions. Forecasting technology is hard, informed people disagree, and a date means little without a definition. Ask who is speaking, what they gain and what evidence they give.
Does AGI mean the AI is conscious?
Not necessarily. Most capability-based definitions do not require consciousness. Consciousness is a separate question, and there is no agreed scientific test for it in machines.
What is alignment?
Alignment is the challenge of making an AI system pursue the goals and values its users and designers intend. It matters for any capable system, whether or not anyone calls it AGI.
Does the EU AI Act depend on whether something is AGI?
No. The Act regulates AI by risk and use, and includes rules for general-purpose AI models. Whether a system counts as AGI is not the test. Check the official text for exact wording.
When do the EU AI Act's high-risk rules apply?
After a 2026 simplification package, obligations for stand-alone high-risk systems (Annex III) apply from 2 December 2027, and for high-risk AI embedded in regulated products (Annex I) from 2 August 2028. This is not legal advice, so check official sources.
When you have finished
Get your certificate of completion
Type your name and download a certificate for this workbook as a PDF, ready to print or to add to LinkedIn. It is made on your own device, so your name is never sent to us. It is a self-declared certificate, not an accredited qualification.
Learn the language
Key terms
- AGI
- Artificial general intelligence. There is no single definition, but it usually means AI that is broadly capable, roughly matching people across most cognitive tasks.
- Narrow AI
- AI built or trained for one task or a small set of tasks, however well it performs them.
- General AI
- AI that handles a wide range of tasks, including ones it was not specifically prepared for.
- ASI
- Artificial superintelligence: AI that outperforms all humans across a broad range of tasks.
- Levels of AGI
- A 2023 framework from Google DeepMind researchers that rates systems by performance, generality and autonomy.
- Alignment
- The challenge of making an AI system pursue the goals and values its users and designers intend.
6 of the workbook's 12 terms. The complete glossary is in the workbook.
Follow the evidence
Sources and checks
Facts last checked: .
These workbooks use AI assistance. See how the workbooks are made.
- Levels of AGI for Operationalizing Progress on the Path to AGIMorris et al., Google DeepMind (arXiv 2311.02462)
- AI Act (includes the application timeline)European Commission, Shaping Europe's digital future
Created by Mickarle Wagstaff-Irons - Micky Irons with the Mickai team. Published by Mickai LTD. Last updated 25 September 2026.
NextKeep going
Where to go next
Recommended for you
What is an AI agent? Agentic AI explained
A plain-English guide to AI agents: how they differ from chatbots and workflows, the parts and the loop, levels of autonomy, the main risks and how to keep a human in control.
Recommended for you
Reasoning models and reasoning engines
What reasoning models are, when extra thinking pays off, how to control it, why a reasoning trace is not proof, and how classical reasoning engines differ.
Recommended for you
How to evaluate a new frontier model the day it lands
A ten-step method for judging a new model: primary sources, licence, architecture, hardware, benchmarks, your own tests, safety, jurisdiction and a decision record.