Silicon Valley GirlAI-First Playbook: Do a Team's Work With AI (2026) | Peter Yang
At a glance
WHAT IT’S REALLY ABOUT
Peter Yang’s AI-first workflow replaces a full content operations team
- Peter describes turning repetitive creator work (newsletter editing, cross-platform posting, podcast prep, and sponsor management) into reusable “skills” so AI can execute end-to-end workflows with minimal copy-paste.
- He explains a practical “self-improving” approach: after each back-and-forth, ask the AI to update the underlying skill text file, then review changes so next runs need fewer corrections.
- A key lever is moving from ChatGPT/Claude chat to Codex/Claude Code because these tools can integrate with APIs, browse when APIs don’t exist, and directly update artifacts like Google Docs.
- He outlines a 5-layer AI adoption framework—from everyday Q&A to agentic automation—and emphasizes building a lightweight “personal OS” folder of skills tied to your principles and business context.
- The conversation also surfaces downsides and cultural shifts: AI can make people feel lazier, encourages low-effort viral “slop,” and raises concerns about kids skipping fundamentals—prompting a focus on values, craft, and learning-by-building.
IDEAS WORTH REMEMBERING
5 ideasTreat “skills” as your real product, not prompts.
Peter frames skills as simple instruction text files that encapsulate a workflow (e.g., edit newsletter, repurpose into social threads, post across platforms). Once the skill exists, you iterate it like code rather than reinventing prompts each time.
Self-improvement is mostly a review loop, not magic autonomy.
After a session where the AI misses the mark, he asks it to update the skill based on the conversation, then he reviews the edits. Over time, the workflow moves closer to “one-shot,” but still needs human oversight.
Leveling up often means switching tools, not learning new tricks.
He recommends moving from standard chat interfaces to Codex/Claude Code because they can run integrations, call APIs, do browser-use, and directly update documents—eliminating the copy/paste tax that keeps many people stuck at Level 2.
Build a “personal OS” by canceling a day of meetings and dumping workflows.
His practical method is to reflect on the past week’s manual work, brain-dump the steps (often via voice tools like Whisper/SuperWhisper), and translate them into modular skills that chain together into full pipelines.
Automate distribution, but don’t automate taste.
Peter stresses the “last 10%” must be human to avoid slop: AI can clean a voice brain-dump into polished copy, but you still set direction, apply judgment, and do final edits to preserve voice and quality.
WORDS WORTH SAVING
5 quotesI fear that I'm getting dumber and, and lazier. Like, if I'm on a flight and there's no, like, internet connection, like, that... I just don't feel like working anymore
— Peter Yang
Like, it's, it's actually important, like I always emphasize like the last 10%, you gotta add your human touch to it. You can't, can't just like AI slopify everything.
— Peter Yang
I, I do think it's a problem 'cause I, I see a lot of creators like just making slop. Like they, they like, they get AI to generate like 10 posts per hour, and like the slop goes vi- viral, and it's kind of like, um, dis- disheartening to see that.
— Peter Yang
Switch from consumption mode to actual build mode.
— Peter Yang
And, uh, unlike like a em- employee, it's never gonna leave you. It's gonna be there.
— Peter Yang
High quality AI-generated summary created from speaker-labeled transcript.
