Just came across this and it made me laugh. Someone built a Claude Code plugin that rewrites all of Claude's responses in caveman-speak. "Why use many token when few token do trick." The actual result? It cuts output tokens by about 75% while keeping all the technical detail intact. There's a companion tool that compresses your memory files too, saving another 45% on input tokens. - Before: "The reason your React component is re-rendering is likely because you're creating a new object reference on each render cycle. When you pass an inline object as a prop, React's shallow comparison sees it as a different object every time, which triggers a re-render. I'd recommend using useMemo to memoize the object." - After: "New object ref each render. Inline object prop = new ref = re-render. Wrap in useMemo." Same fix. Quarter of the words. It's already got 5.5k stars on GitHub in 3 days, so clearly it's struck a nerve. One-line install if you want to try it. Not for me personally (I like my Claude verbose and chatty), but if you're burning through tokens on tasks, this is a funny and genuinely useful hack. Link: https://github.com/JuliusBrussee/caveman

I saw Sabrina Ramanov's video about opencode which is open source ai coding agent. She said its built for privacy. From time to time, I hit my usage limit on claude code and this one might be helpful for me during those times and just wondering if anyone used it before. https://www.youtube.com/shorts/0WPhKD-L9-Q https://opencode.ai/

Anyone else watching NASA live feed on YouTube? 🤓

Found this in a computer magazine over the weekend. Apparently "Microslop" is now the unofficial name for Copilot, and Microsoft is so bothered by it they've been banning the word on their own Discord server. Which obviously made it more popular. If you're on a Windows machine you've probably noticed it creeping into everything uninvited. Thought this might give you a laugh.

Anyone used Gemma 4 to create Ai apps yet?