Cindy Zhu.
Free guides
Start here Claude

The one-hour Claude watchlist

Hey, it's Cindy ๐ŸŒฑ You commented TONIGHT, so here is the actual watchlist: four lanes, real videos with real run times.
btw the waitlist for AI System Club is now open! If you're keen on using AI to automate your content, video editing, job search, business or build AI agents, feel free to sign up (only if you want to hehe). Ps. the waitlist gets first dibs:)
๐Ÿ’ก The rule that makes this work. Watching a Claude video and understanding a Claude video feel identical while you are on the sofa. They are not the same, and you find out the difference the first time you open a blank chat. So the hour is split three ways: 20 minutes watching, 20 minutes copying the exact thing they did, 20 minutes pointing it at your own work. The third block is where the learning actually happens.

the shape of the hour

How the 60 minutes are spent

0:00 to 0:20, watch One build video, start to finish, phone face down. Not a "top 10 AI tools" video. Someone building one specific thing.
0:20 to 0:40, copy Rewatch with Claude open beside it. Pause, do the step yourself, unpause. You are copying their example on purpose, not inventing yet.
0:40 to 0:55, transplant Same workflow, your own inputs. Your inbox, your content plan, your admin. This is where it stops being a tutorial.
0:55 to 1:00, write it down Five lines on what you built and what broke. This is what makes tomorrow's hour twice as useful as tonight's.

pick one

The four lanes

Choose by what you want out the other side, not by which video looks most impressive. Every video below is free, on YouTube, and I have listed the real run time so the hour actually fits.

Lane 1: you want Claude to do your everyday work

Watch: Claude Cowork: 8 Use Cases Every Marketer Should Know by Grace Leung, 20 minutes 51 seconds. Eight worked examples of Claude doing real work inside real files, no code, no terminal. It is framed for marketers but the shape of it transfers to any desk job.

Copy: pick the single use case closest to something you did last week and rebuild it exactly as shown, using her example first.

Transplant: run the same use case on your own file. Her channel: youtube.com/@graceleungyl.

Lane 2: you want Claude to make you money

Watch: How I'd Make Money with Claude if my life depended on it by Nate Herk, 9 minutes 19 seconds. Short and unusually concrete about what is actually sellable. Because it is only nine minutes, follow it with the first ten minutes of How to Build a One Person AI Business, 24 minutes 52 seconds in full, and come back for the rest another night.

Copy: build the smallest version of one offer he describes, end to end, badly. Badly is fine. Finished is the point.

Transplant: rebuild it for a business you actually know, ideally one you have worked in. His channel: youtube.com/@nateherk.

Lane 3: you want Claude to have a team of skills

Watch: Claude Skills: Build Your First AI Marketing Team in 16 Minutes by Grace Leung, 16 minutes 42 seconds. Skills are the single highest-leverage Claude feature for non-developers, because they turn a prompt you keep re-explaining into a command you type once. If you want the bigger version afterwards, Claude Code: Build Your Full AI Marketing Team runs 17 minutes 35 seconds.

Copy: build the exact skill she builds, install it, run it once on her example.

Transplant: rewrite that skill's instructions for a task you repeat weekly, then run it on this week's version.

Lane 4: you already build, you want depth

Watch: Every Claude Code Skill I Use to Drive My Entire Development Process by Cole Medin, 23 minutes exactly. This one assumes you are comfortable in a terminal and is deliberately last, because it makes far more sense after you have shipped something small.

Copy: take one skill from his setup and reproduce it in your own project.

Transplant: replace one manual step of your real workflow with it. His channel: youtube.com/@ColeMedin.

๐ŸŒฑ Genuinely, pick one. Four lanes exist so you find yours, not so you watch four hours. The person who finishes lane 1 tonight is ahead of the person who bookmarks all four and opens none.

the honest correction

About "just watch Anthropic's channel"

I said in the reel that Anthropic's own channel is underrated, and I want to be accurate about what is actually on it. Anthropic's YouTube channel is mostly short announcement and research videos, often one to two minutes, like Introducing Cowork or Let Claude handle work in your browser. They are excellent for knowing what shipped. They are not hour-long walkthroughs, so do not plan your evening around them.

The actual Anthropic-made teaching lives somewhere else, and it is free:

  1. ๐ŸŽ“ Anthropic Academy. Free self-paced courses with completion certificates, at academy.claude.com. You register with an email and no payment method is requested anywhere in the flow.
  2. ๐Ÿ“˜ Start with Claude 101. academy.claude.com/courses/claude-101 is the no-code starting point and the honest answer to "where do I begin". Roughly the length of the watch block, so it slots straight into lane 1.
  3. ๐Ÿงฉ Then Introduction to Agent Skills. Same catalogue. Take it after you have built one skill by hand from lane 3, because the concepts land far better once you have already broken one.
  4. ๐Ÿ”” Keep the channel for news only. Subscribe to the channel so you hear about new features in a minute rather than through a reel three weeks later. That is the job it is genuinely good at.

the copy block

What to actually type in the middle 20 minutes

The copy block fails when you sit there thinking "right, so what do I ask it". Use these two prompts to bridge from the video into your own work. Open claude.ai in a second window before you press play.

First, while the video is still fresh, get the workflow out of your head and onto the page:

๐Ÿง  turn the video into a workflow
I just watched a tutorial and I want to rebuild what they did rather than only understand it. Here is what happened in the video, in my own messy words: [describe what they built, what they typed, and what came out, even if you are fuzzy on the details].

Turn that into a numbered workflow I can follow right now, step by step, with the exact thing I should type at each step written out in full rather than described. Where I have clearly missed or misremembered a step, fill the gap and mark it as your assumption so I can check it against the video. Where a step depends on a tool or setting I might not have switched on, tell me the setting and where to find it before I get stuck halfway.

Keep it to the smallest version that still produces a finished result. I would rather build a small thing tonight than half of an impressive one.

Then, in the last block, move it onto your own material:

๐ŸŒฑ transplant it onto my work
Now run that same workflow on my actual work instead of the tutorial's example. Here is my situation: [your role, the task you repeat most often, and the tool or file it usually lives in]. Here is a real example of the input I would give it: [paste something real, an email, a plan, a spreadsheet row, a brief].

Adapt the workflow to my inputs and tell me clearly where my version has to differ from the tutorial's version, and why. If any step does not transfer, say so plainly rather than inventing a workaround that will fall over the second time I run it.

Finish by telling me which single step of this I should turn into a reusable saved prompt or a skill, and write that saved version out for me in full so I can keep it.
๐Ÿ’ก Pause discipline. The temptation is to watch the whole video and then rebuild from memory. Do it the other way: pause on each step, do the step, unpause. It feels slower and it is roughly twice as fast, because you never build on top of a step you got wrong.

when you want more

Two things for after tonight

The weekend version Nick Saraev's Claude Code Marketing Full Course, 6 hours, free. Not a tonight video. A genuinely good Saturday, once you have built something small first.
The wider map If one hour turns into a weekend, I wrote the 9-step weekend playbook, and the gentler entry point is start here, 4 free things.
๐ŸŒฑ Why one hour is enough to matter. Not because an hour makes you good. Because an hour of building moves you out of the group who have opinions about AI and into the group who have used it on their own work. That is a smaller group than you would expect, and the gap between the two grows every month.

all the links

The whole watchlist in one place

Want a path instead of a playlist? Inside AI System Club I teach people to learn AI from scratch and build real systems, in order, with support. Doors open soon and founding pricing dies at close. ๐ŸŒฑ Join the waitlist (link in my bio too).
the waitlist is now open

AI System Club is launching soon

Build AI systems with me through recorded lessons you can follow at your own pace. Join the waitlist for first access and the founding price.

btw this list gets first dibs
the moment doors open :)