Schemas and Propositions
Subscribe
Sign in
Home
Notes
Archive
About
We've been using AI wrong: Agent Skills and the Semantic Powers of LLMs
How a folder with some text files can move your use of AI from chatting with a bot to getting things done and what you can learn about how LLMs work in…
Feb 2
•
Dominik Lukeš
3
December 2025
How LLMs Understand Text (by me) and how humans don't (by Kahneman)
Summary of a talk I gave at Kellogg College, University of Oxford, on November 17, 2025 with some related quotes from Daniel Kahneman.
Dec 7, 2025
•
Dominik Lukeš
2
2
September 2025
Symbols and Manipulation
In which we examine how easy it is to be completely wrong about the nature of the mind by confusing symbols in Principia Mathematica for symbols in the…
Sep 8, 2025
•
Dominik Lukeš
2
August 2025
Karpathy Thinks in Code but a Computer Doesn't
In which we investigate what humans and LLMs have in common and what sets us apart from computers: schemas. But in which we also have some fun with…
Aug 31, 2025
•
Dominik Lukeš
1
50 First Dates and the ChatGPT Memory Problem
In which we learn how Large Language Models actually generate their responses and why everybody is so obsessed with context engineering now.
Aug 27, 2025
•
Dominik Lukeš
1
1
Why Schemas Matter
And why you're probably wrong about how the mind works.
Aug 22, 2025
•
Dominik Lukeš
1
This site requires JavaScript to run correctly. Please
turn on JavaScript
or unblock scripts
