My 8000-word note on agents: Covering:. 1. An overview of agents. 2. How the capability of an AI-powered agent is determined by the set of tools it has access to and its capability for planning. 3. How to select the best set of tools for your agent. 4.

$100 for anyone who can show me how to get ChatGPT to stop using emdashes. it's driving me insane

13 years. 6 books. All metrics are flawed, but it still makes me happy to see this. I love writing, and I hope to improve at it over time.

Manus AI Manus AI chose to focus on context engineering rather than developing models. If you were to start an agentic company today, which would you invest in?.

3. Dynamic few shot prompting. They cautioned against using the traditional few shot prompting for agents. Seeing the same few examples again and again will cause the agent to overfit to these examples. Ex: if you ask the agent to process a batch of 20 resumes, and one example

2. Tool use. Given how easy it is to add new tools (e.g., with MCP servers), the number of tools a user adds to an agent can explode. Too many tools make it easier for the agent to choose the wrong action, making them dumber. They caution against removing tools mid-iteration.

Very useful tips on tool use and memory from Manus's context engineering blog post. Key takeaways. 1. Reversible compact summary. Most models allow 128K context, which can easily fill up after a few turns when working with data like PDFs or web pages. When the context gets

I’m slowly beginning to accept that my productivity, when working with AI coding agents, is limited by my human brain. AI can do many tasks in parallel, but I can only track the context of a few, so I only run a few tasks at a time. I am the bottleneck.

4. Sniffly also allows me to walk through all my previous instructions and model’s responses, and I can also share them with my collaborators if needed. Sniffly is open-sourced and can be used without installation. uvx sniffly@latest init. GitHub:

3. While most of the time, Claude Code can only go up to 10 steps before I need to interrupt it, it can occasionally go close to 100 steps. Just a year ago, people told me it was hard to get an agent to go above 5 steps!. Claude Code’s favorite tools are, unsurprisingly, search

2. Traditional metrics of engineering hours/days don’t work for AI. Two metrics I use to evaluate the complexity of a project:. - how many instructions I need to give AI until it completes a project.- how often I have to interrupt it because it goes into the wrong direction

I open sourced Sniffly, a tool that analyzes Claude Code logs to help me understand my usage patterns and errors. Key learnings. 1. The biggest type of errors Claude Code made is Content Not Found (20 - 30%). It tries to find files or functions that don't exist. So I

RT @miramurati: Thinking Machines Lab exists to empower humanity through advancing collaborative general intelligence. We're building mult….

I asked Claude Code to fix my bug and it just refused lol. "Your app is working fine. This is a minor issue that doesn't break core functionality."

I think I might've accidentally jailbreak-ed Claude. I asked Claude to craft a conversation between two characters for a story I'm writing and they suddenly started reciting Claude's system instructions.

omg thank you jensen.

Finally!!! An AI frontier lab run by adults. Congrats to @miramurati, @johnschulman2, @barret_zoph, @Luke_Metz, and the rest of the stellar team!. Fingers crossed for less AGI and more cool products 🙏.

RT @karinanguyen_: Highly recommend this book by @chipro esp. for product designers / software engineers who want to learn about model trai….

Wrote a quick note (with examples) of common pitfalls that I’ve seen, both from public case studies and from my personal experience. Would love to hear from your experience about the pitfalls you've seen!