The Explainer Series
WHAT IS?
One term per page. Defined precisely in the first paragraph, then unpacked: how it works, what it gets confused with, what it can’t do, and when to use it.
All entries
5 live · 6 comingWhat Is AGI?
Artificial general intelligence is the goal AI was founded to reach and the term most abused in AI marketing. Here's the precise definition, how it differs from today's AI, and why there is still no agreed finish line.
- AGI means AI that can learn and perform across essentially any intellectual task a capable human can, rather than excelling at one bounded task. Generality, not raw skill at a single benchmark, is the defining property.
- There is no agreed definition or test for AGI. Labs, researchers, and contracts each define it differently, so any claim that 'we have reached AGI' needs the definition attached before it means anything.
What Is SI (Superintelligence)?
Superintelligence is intelligence that substantially exceeds the best human minds across most domains. It does not exist, and everything said about it is a forecast. Here's the precise definition, the arguments for and against, and the part that actually matters today.
Read entry
What Is the Internet?
Not a cloud floating in space — a physical network of cables, satellites, and routers moving data in packets across the globe. Here's what the internet actually is, underneath the everyday habit of using it.
Read entry
What Is Compute?
Compute is the active engine of the digital universe — the transformation layer, sitting apart from storage (keeping data still) and networking (moving data around). Here's what it actually is, how it works, and where it hits hard physical limits.
Read entryWhat Is AI?
Not a single machine or invention — a field of computer science for building systems that learn patterns from data and use them to do things that used to require human judgment. Here's the precise version, without the hype or the dismissiveness.
Read entry
How every entry works
No listicles. No ten-tab research project.- 01
Definition first
The answer sits in the first paragraph, stated precisely, before anything else.
- 02
How it works
A working mental model of the mechanism, not a glossary of jargon.
- 03
What it is not
Side-by-side with the neighboring terms it is constantly confused with.
- 04
Limits and use
Honest limitations, and when the thing is actually the right tool.
Coming next
Same format, same core topics- AIWhat Is Machine Learning?The method underneath most of modern AI, explained without the math.
- AIWhat Is a Large Language Model?How next-token prediction turns into conversation, code, and everything in between.
- AIWhat Is Generative AI?Prediction vs. generation, and why one inverts the other.
- EconomicsWhat Is Inflation?
- BusinessWhat Is Product-Market Fit?
- TechnologyWhat Is an API?