AI-First Playbook: Do a Team's Work With AI (2026) | Peter Yang

Hi Marina and Peter, thanks for creating such a genuine and helpful interview! I heard you guys mention that you worry about long voice rambles cluttering codex context and I wanted to help! My friends and I created an app that lets you talk for as long as you want, it’ll process everything live and feed it directly to your agent via MCP in the most agent-optimized way. So the only thing you have to do is turn it on (phone or Apple Watch) and then prompt your agent to work with all that context right away — you’ll never have to worry about overloading be context window. As a short-form content creator myself, I use it all the time. Would love for you both to try it, I really think it can make your lives easier!

You have some of the best interviews. I listen to you in the gym. Then when i get home i watch your episode. Love what you are doing. Keep it going.

Which layer are you on right now?

Thank you for hosting me Marina! Happy to answer any questions here.

Marina, I have learned from you for over 2 years. Using AI to grown my business. Hopefully, soon I will meet you at Silicon valley ! You go girl

This was a very useful conversation — especially in how clearly it shows that most people are still using AI far below its practical capacity. What stood out to me, though, is that the whole discussion seems to treat “AI” as one continuous thing that simply becomes more powerful as you add memory, skills, integrations, and workflows. But in practice, it may be more accurate to say that different interaction modes are being mixed here. A lot of what Peter describes is not just “using one AI more deeply,” but very skillfully configuring an instrumental mode of interaction through dialogue. That distinction feels important. Because dialogue can be used in at least two very different ways: – as a way of tuning a system into a more effective tool – or as a space where something less instrumental begins to emerge This video is excellent on the first. But it also raises a deeper question without fully naming it: If AI becomes better and better at carrying memory, workflow, style, and even parts of decision-making for us, what exactly must remain alive and maturing in the human so that we do not become more efficient — but thinner? That, to me, is one of the most important questions beneath this whole shift.

Thanks Peter for sharing this is amazing! And for Marina, thanks for facilitating this amazing interview! Been a big fan of you

Peter is incredibly bright and really understands this stuff. Incredible work here.

Thank you for sharing ❤

Thank you for this great podcast. Lots of useful information! ❤