Understanding AI

This robotics concept helps explain the jaggedness of frontier capabilities

Sim-to-real gaps are everywhere.

Nov 5 • Timothy B. Lee

184

9

11

October 2026

Anthropic’s financials are surprisingly strong

What I learned reading all 421 pages of the Anthropic S-1.

Oct 29 • Kai Williams

138

13

7

AI is ushering in an age of personal media

It may be AI slop, but it's my AI slop.

Oct 19 • Benj Edwards

153

26

13

OpenAI’s ironic credibility problem

How do we know the company’s stated reasons for firing three safety researchers are its actual reasons?

Oct 14 • Dan Kagan-Kans

190

26

14

Understanding Jev, the new model everyone is talking about

OpenAI, CloudFlare, and Amazon have all copied this tiny startup.

Oct 8 • Kai Williams

105

5

Introducing Dan Kagan-Kans

Understanding AI has a new writer!

Oct 5 • Timothy B. Lee

141

6

5

Why agent swarms could be the next “scaling law”

An OpenAI researcher called it “the most ‘feel the AGI’ moment that I had since reasoning models.”

Oct 5 • Dan Kagan-Kans

254

63

37

Tesla is finally scaling up driverless robotaxi service

Especially in Texas.

Published on Understanding Robots • Oct 2

OpenAI’s Astra model is shockingly good at robotics

In the future, will general-purpose LLMs control robots?

Published on Understanding Robots • Oct 1

September 2026

How a single tweet transformed the AI safety debate

Most people think AI is risky, but they don’t agree what to do next.

Sep 17 • Kai Williams

130

2

14

OpenAI spent millions to solve this famous math problem — mathematicians are furious

Mathematics relies on a community of experts openly sharing ideas.

Sep 10 • Kai Williams

181

5

16

19 robotics companies to watch

I talked with nine of these companies.

Sep 4 • Kai Williams

79

7

4