Nate talks about using Plan Mode within Claude Code all the time. The basic idea is simple: before Claude starts editing files or writing code, you have it slow down and create a plan first. In Plan Mode, Claude focuses on understanding the request, reviewing the codebase, identifying the right files, and outlining the implementation approach before making changes. On bigger tasks it can save tokens overall by helping Claude avoid bad assumptions, wrong edits, and rework. It helps prevent the classic AI coding problem of: “It started building before it fully understood what I wanted.” In Claude Code, press Shift + Tab to enter Plan Mode. Then describe what you want done and ask Claude to create a step-by-step plan first. Review the plan, correct anything that looks off, and only then let it move into implementation. the implementation approach before making changes. Bonus Tip : Use Plan Mode in conjunction with the Grill Me skill, as discussed by Nate yesterday : https://www.youtube.com/watch?v=c0kaKxM2pHg

Thank you. Now I’m going to give them something to say🤓

I recently started learning in Classroom curriculum and the 7-Day AI Challenge, but I'm feeling a bit lost about the best learning path. I'm unsure whether I should focus on building things right away or spend more time understanding the fundamentals first. For example, would it make sense to jump straight into creating AI workflows using tools like Claude Code and learn by doing, even if I don't fully understand what's happening under the hood? Or would it be better to first learn the basics of tools like n8n, APIs, different endpoint types, workflow architecture, and how AI systems connect together before building more complex projects? I'm trying to find the right balance between practical experience and foundational knowledge. For those who have already gone through this learning process, what approach worked best for you? What do you wish you had learned earlier, and what did you learn effectively just by building and making mistakes?

No, not the business type. The personality trait. I wonder if this is a new personality trait business will hire on in this age of AI.

I built my AI Operating system but when I wanted to connect HoneyBook (my CRM) to my operating system, I found out that HoneyBook doesn’t offer APIs or MCPs. It’s a bummer because that’s where all my client information is located! What’s the workaround here? FYI, I’m a wedding photographer and videographer.