I’ve always been opinionated about how software should work. Mainly, it should be fast. The bounds of it should be “knowable.” The contract you have with it should be “sane” (i.e., you just own it).
The rest of Craig’s post is about how he used Claude Code to fulfill his wildest software dreams. (In his own words, he’s gone “software bonkers.”)
I’m not going to lie, the tools he made sound pretty cool. But…
Are the bounds of Claude Code knowable? Is the contract you have with it “sane?” Do you just own it?
In some ways, these are genuine questions. I don’t know much about Claude Code. Maybe it isn’t just another product that will twine itself around your digital life. Maybe its omnipotence is beneficial and it really can help you make fast, sane, knowable tools. If so, cool. But I remain skeptical.
Christian Leithart 