Join 📚 Josh Beckman's Highlights
A batch of the best highlights from what Josh's read, .
I'm continually impressed with Claude's ability to create its own benchmarking systems and rip through a series of hypotheses about how to achieve some hard-to-define optimal outcome.
I recommend trying this when you're deep into a problem looking for an optimal path: give the agent a laboratory.
Give Your Agent a Laboratory, Pt. II
Writing
Interestingly, Ojisan TCG didn’t start as a competitive game. The first set of cards was designed purely for collecting, but the local children quickly turned it into something more dynamic. They began comparing stats and declaring, “My card is stronger than yours.”
Seeing this, the game’s creator decided to take it to the next level. New rules were introduced, allowing the cards to be used in actual battles. The objective isn’t to defeat the opponent’s card but to outplay it based on the characters’ skills and abilities.
Middle-Aged Man Trading Cards Go Viral in Rural Japan Town
Ynes Sarah Filleul
Elisa Baniassad and Alexander Summers have this great paper [Reframing the Liskov Substitution Principle through the Lens of Testing](https://www.cs.ubc.ca/~alexsumm/papers/BaniassadSummers21.pdf) , where they teach LSP as “the superclasses test suite should automatically be runnable, and pass, on the child class.” Go read it, it’s great.
When to Prefer Inheritance to Composition
Hillel Wayne
...catch up on these, and many more highlights
Readwise