A section of Craig Mod's newsletter, Roden discussing how he's using Claude Code to develop things (like The Good Place, a "reverse-chron-text-feed" clone
with specific rules).
It’s no exaggeration to say that using Claude Code to build The Good Place (and also a bunch of other small tools and projects) is one of the most astonishing computing experiences of my life.
It’s difficult to articulate how utterly empowering a tool like Claude Code (paired with malleable software, open software, open systems (i.e., not iOS/iPadOS)) is for someone like me: someone with a strong technical background who can guide the LLM, knows which questions to ask, and knows how to keep it from going off on weird tangents.
I routinely test out dumb little software experiments, learn about new libraries by watching Claude think through problems, and am increasingly in a state of idiotic rapturous delight by how quickly I can go from notion to function.
Required Caveat: There are many moral and ethical issues with using LLMs, but building software feels like a) one of the few truly ethically “clean”(er) uses (trained on open source code, etc.), and b) maybe the thing LLMs most excel at (science, medicine, and code seems to be their wheelhouse in terms of energy-in to manifold-value-out). Even with that caveat, I realize many of you will still be driven to madness by my breathless love fest here, but this technology is barreling forward, and if I’m going to criticize it (and I’m happy to, delighted to), I want to know the precise contours of its delights and horrors.