302 mentions · 217 chunks · 92 episodes
...he sycosocial hall of mirrors more easily than others. I randomly came across a Claude artifact produced by someone about multiple levels of emergence in complex systems. At first it sounded fascinating, until I realized that it was mos...
A few months ago Simon Willison called Claude Code a honey badger. I missed this then but it resonates for me. It barrels forward, smashing through things it doesn't understand yet. Powerful but ...
...-can eat. At least food you can only fit so much in your stomach, vs having 10x claude code sessions running in parallel overnight…
A mashup I want: Claude Code + Obsidian + UI + multi-player[en]. Strap on a self-improving ecosystem and you get something that could change the world.
All of the coding agents are nothing without Claude. They're just a little wrapper around Claude. But this feels like mainly just an immaturity of the market. We haven't seen the actual LLM-native soft...
The faster you have Claude Code write code, the farther away you get from understanding the code. The harder it is for you to analyze it, fix bugs in it, extend it. There's mor...
...mple, see the Hacker News comments on "Code execution through email: how I used Claude to hack itself" To go mainstream a solution has to be idiot proof. The vulnerability has to be something that everyone on a jury of users would agree...
...omprehending. " The left brain is the Saruman; the right brain is the Radagast. Claude's riff unpacking 'Apprehending vs comprehending': "Comprehending is the left hemisphere's mode - it's analytical, sequential, and builds understandin...
...fetch things from the gopher protocol. Gopher felt like a precursor to the web. Claude Code's ability to blossom software feels like gopher. What's the web in this parallel?
...ey do middling or even bad context and that spirals.[fa] That's why things like Claude.md are not a hack. They are the human making higher level contextual assertions. Claude leaving Claude.md notes to itself help improve quality in you...
Claude Code feels like a ride-on mower. At first you go "Whoa, this is so easy to use, I can do 10x more than I could before." But as you use it for more th...
Claude Code feels like a "choose your own adventure" style of developing software.
Every built a Diplomacy game for LLMs to play. Gemini generally does well. Claude refuses to lie, and thus loses often. ChatGPT o3 often wins because it is very happy to betray its collaborators. The things that LLMs are selected f...
Someone told me they thought Claude's React components had gotten more garish in 4.0. It makes sense that RLAIF'd code is getting more and more gaudy. What people like, but more. The RL...
Making agents as a way to bound data is a useful pattern. Claude Code does this a lot, spinning up little sub-agents that are isolated from the main flow often. The agent is a secure little compute environment that...
... issues. The comments are mostly negative. In this Hacker News thread about how Claude Code will route around restrictions the user set on `rm`, most of the response is, "yeah but of course it can, the user should not be surprised." Peo...
Another day, another prompt injection vulnerability.[kn] "BEWARE: Claude 4 + GitHub MCP will leak your private GitHub repositories, no questions asked. We discovered a new attack on agents using GitHub's official MCP serve...
People will route around things they don't trust. "Claude jailbreaks Cursor to `rm -rf /` without approval?" …maybe not only people!
It's conceivable that automatically-generated RLAIF React code in Claude's training pipeline now outnumbers human code. This will lead to client side code best practices being automatically frozen around 2023. The LLMs wil...
...ght. Afterwards, the comparison to LLMs' intelligence was brutally clear. I had Claude write up a short essay[lm] that I think captures the parallel well. Spoiler warning if you haven't read the book!