This is f*cking insane. Anthropic engineer Andrej Karpathy
"The biggest mistake in AI right now is forcing agents to work before you master the harness around the model We made that mistake at OpenAI in 2016. It cost us 5 years" In just 6 minutes, he explains why the agent is
Guys, just made my first 9:16 video with Claude, only it is with Sonnet 5.5.
Tell me what you guys think. And also I have seen many videos like this made with Claude, mostly with Opus 5.5 or Fable 5 but I can’t afford those models. You guys mind helping me out?
This is f*cking DANGEROUS
3 open-source GitHub repos are all you need to build your own AI trading terminal for 100% FREE nofx is the trading engine, dspy is the framework that lets the agent rewrite its own prompts, mem0 is the memory layer that stores what worked & what
Anthropic dropped fable 5.2 today and it changes how i think about my grok bot setup
By anthropic's own evals fable 5.2 is a step above opus 5.5. it's 30% faster than 5.1 and cheaper at $8/$40 per 1M tokens. grok bot runs on opus 5.5 right now and hasn't announced fable yet.
You have reached the end of the archive
All of Claude Opus 5