Big Idea

The argument in one line.

Prompting an AI chatbot is already the old model — the builders of these tools have moved to loop engineering, where you set a goal once and agents handle discovery, parallel execution, verification, and iteration without you in the chain.

Who This Is For

Read if. Skip if.

READ IF YOU ARE…

You use Claude Code, Codex, or any agentic tool and want more output without manually prompting every step.
You have weekly or monthly repeatable tasks — client updates, content research, product copy — that you still do manually.
You want to understand how senior AI builders actually work, not just how they post about it.
You have heard the word agents but every explanation has been too abstract or too technical to act on.

SKIP IF…

You need production-grade architecture — this is conceptual plus light demo, not engineering-depth.
You already build multi-agent pipelines professionally; the frameworks shown are intentionally beginner-accessible.

TL;DR

The full version, fast.

The dominant prompting pattern — human types, AI outputs, human reads, repeat — is being replaced by loop engineering: you set a goal once, agents handle discovery, planning, parallel execution, and verification, then iterate until a loop condition is met. The video contrasts open loops (expensive, wide search space, no budget ceiling) with closed loops (bounded goal, human-defined path, normal budget) and recommends closed loops for almost everyone. A live demo builds a 3-agent pickleball e-commerce system in Claude Code — one agent builds a personality quiz, one scouts content opportunities on Reddit and competitor sites, one turns research into launch marketing. An orchestrator agent then manages the whole fleet on a weekly schedule using an external memory file to carry state across cycles.

Free for members

Chat with this breakdown — free.

Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.

Create a free account →

Chapters

Where the time goes.

00:00 – 00:43

01 · Why the top AI programmers stopped prompting

Cold open establishes credibility via bcherny and steipete quotes — I don't prompt Claude anymore, I have loops running.

00:44 – 01:27

02 · What a loop actually is

Side-by-side: old way (human drives every step) vs. new way (goal set once, agents loop). Visual diagram introduced.

01:28 – 02:44

03 · The 5 stages of a loop

GOAL to DISCOVERY to PLANNING to EXECUTION to VERIFICATION to SHIP or ITERATE. Memory lives outside conversation to persist state. Orchestrator + specialist fleet diagram.

02:45 – 03:34

04 · Open loops vs. closed loops

Open = wide, expensive, unlimited budget only. Closed = bounded goal, budget-safe, recommended. Side-by-side comparison card shown.

03:35 – 05:00

05 · Real-world example: growing a pickleball store

3 tasks mapped: personality quiz (lead magnet), content scout (Reddit/competitors), growth agent (marketing). Goal: grow site automatically without doing every task manually.

05:00 – 07:16

06 · The 3-agent system, mapped out visually

Orchestrator owns goal, delegates to Builder, Scout, Growth Agent. Loop condition: fresh ideas? Lead magnet pipeline full? Weekly scheduled trigger.

07:16 – 09:21

07 · Building it live in Claude Code

Pastes builder, scout, and growth agent prompts into separate Claude Code sessions running in parallel.

11:18 – 13:38

08 · Checking the results

Harry Potter quiz one-shotted and live. Scout ranked 8 content opportunities. Growth agent produced 12 site link placements, launch email, social captions, next quiz recommendation.

13:38 – 15:35

09 · Automating with an orchestrator agent

Orchestrator prompt reads next-steps.md, delegates to all 3 sub-agents, synthesizes into unified action plan, evaluates 3 loop conditions, reruns if not met, scheduled weekly.

15:35 – 18:22

10 · QUIZ: 4 real-world loop scenarios

Freelancer (Friday client updates), Student (Sunday AI briefing), Shop owner (monthly product copy rewrite), YouTube creator (Monday idea scoring). Each has answer card with loop condition and lesson.

18:22 – 19:18

11 · Where AI falls short

Honest caveat: YouTube topic scoring is imperfect. Keep judgment calls in the human layer. Soft subscribe CTA.

Atomic Insights

Lines worth screenshotting.

The creators of Claude Code and OpenClaw both said they stopped prompting and now write loops — the tools themselves changed how their builders work.
An open loop burns massive tokens because the agent roams freely; a closed loop gives it a bounded goal and a budget it cannot escape.
The 5 stages of every loop are fixed: discovery, planning, execution, verification, and iteration — the only thing you set is the goal.
External memory is the primitive that makes loops compound: without a file that persists state across cycles, every run starts from scratch.
Running 3 agents in parallel in Claude Code costs the same as running them sequentially but finishes in a fraction of the time.
A loop condition is just a yes/no question the agent asks itself at the end: did we meet the goal? If not, it keeps going.
The orchestrator pattern separates goal-ownership from execution: the orchestrator reads last-cycle memory, delegates only new work, and synthesizes all outputs.
A skill file — a document the agent loads that knows each client name, preferred tone, and project goal — is what turns a generic loop into a personalized one.
Open loops are only practical with unlimited budget; closed loops are what everyone outside a major tech company should be running.
The honest caveat: AI is not always right about which YouTube topics perform — keep judgment calls in the human layer, not the loop.

Takeaway

How to design a system that works while you are not watching.

WHAT TO LEARN

The shift from prompting to looping is a shift in where you spend your attention — from driving every step to defining the goal and the condition that ends the run.

A loop condition is the most important line in any agent prompt: without a specific yes/no question the agent can evaluate, you get an open-ended run that never knows when to stop.
External memory — a file the agent writes to and reads from between cycles — is what separates a loop that compounds from one that resets every week.
Closed loops are the practical default: define the goal, sketch the path, set the evaluation. Open loops are only viable when token cost is not a constraint.
Running agents in parallel does not cost more than running them sequentially; it just returns results faster — the constraint is your ability to define non-overlapping subtasks.
The orchestrator pattern works because it separates goal-ownership from execution: one agent knows what was done and what remains, the others just execute their bounded task.
A skill file — a document loaded at the start of each cycle with client-specific or context-specific knowledge — is what makes a generic loop produce personalized output.
Keeping the human in the loop matters most at the judgment layer: loops are excellent at research, ranking, and drafting; they are unreliable at predicting what a real audience will actually respond to.

Glossary

Terms worth knowing.

Loop engineering: A prompting paradigm where you define a goal once and an agent system handles discovery, planning, execution, verification, and iteration automatically, rather than you driving each step manually.
Open loop: A loop with no bounded search space — the agent discovers its own tasks and continues until it runs out of ideas or budget, which can be very expensive.
Closed loop: A loop with a defined goal, a human-built path, and a clear evaluation at each step — keeps costs predictable and output focused.
Loop condition: A yes/no question the agent evaluates at the end of each cycle (e.g., Are there 3 unacted content ideas?) that determines whether to ship or iterate.
Orchestrator agent: A top-level agent that owns the goal, reads memory from the previous cycle, delegates subtasks to specialist agents, and synthesizes their outputs into a unified action plan.
Memory file: A file (e.g., next-steps.md) that lives outside the conversation context and persists state across loop cycles, allowing agents to know what was completed and what still needs work.
Skill file: A document an agent loads at the start of a loop that encodes persistent context — client names, preferred tones, project goals — so output is personalized rather than generic.
Claude Code: Anthropic's agentic coding tool, available on Claude Pro and above, that can run multiple parallel agent sessions and interact with files, terminals, and the web.

Resources

Things they pointed at.

00:00toolClaude Code ↗

00:12toolOpenClaw

07:43linkLoop Engineering Prompts (Notion) ↗

Quotables

Lines you could clip.

00:12

“I don't prompt Claude anymore. I have loops running that prompt Claude and figuring out what to do. My job is to write loops.”

Quotable authority statement, strong standalone claim→ TikTok hook↗ Tweet quote

01:24

“This is slow, and there are better ways now.”

Clean pivot line at diagram transition→ IG reel cold open↗ Tweet quote

16:06

“Automate the admin wrapper around your creative work first. That's where the hours disappear.”

Actionable one-liner with clear user pain→ TikTok hook↗ Tweet quote

The Script

Word for word.

Read-along

Don't just watch it. Burn it in.

See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.

analogy

00:00There's a new idea that the top AI users and programmers in the world are all talking about, loop engineering or sometimes it's simply referred to as loops. In particular, Boris Churney who created ClotCode and Peter Steinberg, who created OpenClaw, have both been talking about this saying that this is now how they program.

00:19They don't just prompt a chatbot, and they definitely don't write code by hand, but instead they create these loops where the AI prompts itself. They're saying that this is the future, but at the same time, it's pretty confusing. So in this video, we're gonna break down exactly what loops are, how to use them, what they're good for, what they're not as good for, and a bunch of real world use cases along with a demo so you can see exactly how to start using this yourself.

00:43So at a high level, here's what a loop looks like. And I should mention that loops are usually talked about with regards to coding, but they can also be used for a ton of other things like content creation or research or teaching yourself a new skill, and we'll cover all of those too. But that being said, they do still all go through these coding tools like clot code or codex, but you don't have to be technical to do this.

01:04So if a terminal window scares you, don't worry. It's totally cool. So to show you this visually, this is what the old way looked like.

01:11You as the human would write a prompt, send it to the agent. It would program something, do something, produce the output, send it back to you.

01:19You read it, write the next prompt, then it produces it again, and you just repeat forever. So you're iterating, but this is slow, and there are better ways now.

01:27So here's what the new method looks like. You start with the goal. This is set by the human once.

01:34Then it goes into discovery. So the agents now find what needs doing.

01:39They plan it out. They break it into clear steps. They execute on it, and this is where they can work in parallel.

01:44You can spin out 15 different agents who each go and do one thing. Then you have the verification step where a verification agent checks, did this accomplish our goal?

01:56If it did, ship it. If it didn't, then it iterates and does it again. Now even when you ship it, if you want, you can add in a step where once it's shipped, an agent looks at it and goes, okay.

02:06We shipped that. What should we do next? And then it iterates again.

02:10The other key piece here is memory, which lives outside of the conversation and keeps track of what is going on, what steps have been completed so that the agents can successfully iterate.

02:20And then another way to look at it that's a little bit more concrete is here is an example workflow. You have the orchestrator agent that overlooks everything.

02:29So it knows what's going on and gives out tasks to the individual agents. So you'd have each of these specialists run their own discovery through ship loop, what we just talked about down here. But then they all come together to create the finished product.

02:45There are two types of loops, an open loop and a closed loop. An open loop is much more broad, and you're basically telling the agent, you go out there, see what we should do, and then just go and do it.

02:57Now, that's really cool because the agent can discover its own things that you wouldn't even think of, but it burns massive tokens because it can go in any direction that it wants. And even once it's done with one direction, it'll go in another, and it'll just keep on going. So say if you work at meta and you have unlimited budget, then this is great.

03:12But otherwise, I would recommend the closed loop, which is where you start with that bounded goal that we talk about. You see the path first to understand what sorts of things that agents might do. There's a clear evaluation at each step, and this keeps budget fairly normal, fairly constrained, so it's not gonna get too expensive.

03:31Obviously, that's gonna vary depending on how much you wanna do with your agent, but it shouldn't be too bad. So with that in mind, I wanna talk about what does this look like in the real world? Not some random one off coding project, but something that a real person would use and have.

03:44So our scenario is you have an online pickleball store where you sell pickleball gear. You have some customers, but not a ton, and you wanna grow it. So we're gonna do a few different tasks to try and accomplish that goal.

03:56First up is we are gonna build one of these BuzzFeed style quizzes that is going to recommend a pickleball paddle based on what your favorite Harry Potter character is. Now that's kinda goofy.

04:07It's kinda silly, but those sorts of things do work and they're pretty shareable. So that's step one. Second is we're gonna build a content research engine that goes out and finds what your potential customers are actually talking about right now.

04:19So looking on social media, looking on blogs, forums, Reddit, etcetera. Seeing what's kind of popular, what are people sharing, what do people wanna know about so that we can go and create content about that on our website.

04:30We are also going to have that looping so it's doing that every single week and consistently getting better over time. And then third, we're gonna have a growth agent that takes all of that stuff, looks at our site, and turns it into ready to use marketing. So I'm talking email campaigns, social captions, site recommendations of where we should link to other things to get more people to take this quiz that we just created, stuff like that.

04:53So that's three agents. It's three jobs all running at the same time. And again, I think people learn best visually, so I'm gonna show you what this looks like visually.

05:00So the goal is grow the pickleball e commerce site automatically without doing every task manually. Now, we have our orchestrator agent. They own that goal.

05:09They read, um, this next steps file. You can kind of ignore this for now, but it's basically telling us what to do as we loop through things so we get better. It delegates and then synthesizes outputs.

05:18Now, the most automatic way to do this is to have an AI agent be the orchestrator. And we will have that. But to start, I'm going to be the one orchestrating the three agents just so you can see it visually.

05:29Because right now, if you have an orchestrator agent in Cloud Code on desktop, it will go and create all of these agents, but it just does it within the one chat. So you don't really get to see what those agents are doing.

05:40Now, it's totally fine. It does the same thing, but I think it's more helpful to learn seeing them happen as like unique tabs open. So I'll show both, but be aware of that.

05:48Then we have agent one is the builder. They're gonna build the Harry Potter quiz that we talked about with email capture at the end. Agent two is the scout.

05:57They're gonna search Reddit, forums, competitor sites, search trends for what pickleball buyers want, what they're talking about. And the loop condition is it's going to loop until we have three or more fresh ideas that we've not yet acted upon. Like, we haven't created content about them, things like that.

06:11And then agent three is our growth agent. It's gonna read the quiz, read the site to understand where it should link people to or try to, like, get people to take the quiz. It'll check which recommendations from the last cycle were implemented and find new opportunities.

06:24So it'll help us create emails like you see here, um, site edits, social captions, next quiz recommendation if we want to, uh, build a new quiz, stuff like that. It'll synthesize all of this. The orchestrator will read these, create one unified action plan, and then start acting on that.

06:42And so the looping comes in. The goal is, is the site growing? So are there fresh content ideas?

06:47Is the lead magnet pipeline full? I mean, are people consistently taking that quiz? And if no, then we'll keep going until we get there.

06:54But if yes, then we'll pause, wait for the next steps. And then we're gonna have these running on a weekly schedule cadence, which will automatically trigger the orchestrator. So like to find new content ideas, to analyze the site and see what we could be doing to, um, get more customers, like emails to send out, stuff like that.

07:10That will happen weekly. All of that feeds back to the orchestrator creating our loop.

07:15Now let's go build it. So we're gonna open Claude, and then we're gonna go to Claude code. Now you will need a pro plan or higher in order to access Claude code, so just be aware of that.

07:25Now we've got our new session here, and then I'll paste in this prompt. I'll link to this as well so you have this. You're the builder.

07:32Your only job is to create a Harry Potter x pickleball personality quiz. The deliverable is a single self contained HTML file saved to slash outputs slash quiz. Six questions total.

07:43Each question should feel fun and slightly absurd. Questions should subtly map to pickleball playing styles, so like aggressive, strategic, social, defensive, etcetera, for possible results. And then we'll have the email capture form that appears before the result is revealed with a CTA that says see your full result.

08:00And then I always like to include this, ask questions if needed, so you can work with the agent to get something that you really like. Then we're gonna hit start on that. Now, while we have that one going, we're gonna hit command n to start a new session and then jump to our second agent.

08:14We're gonna prompt it with this. You are the scout. Job is to research real content opportunities for a pickleball e commerce site completely independent of the quiz being built in parallel.

08:25Go out to the internet and find what pickleball players, buyers, and enthusiasts are actually talking about searching for and struggling with right now. To some research sources, the Reddit, these subreddits, search trends, competitor sites, as well as YouTube.

08:39What videos are getting traction? For each opportunity, score it on audience size, purchase intent, will people actually use this to buy stuff, content gap, is it underserved, and then quiz or lead magnet potential.

08:53Could it become some sort of quiz or tool or guide? Deliverable, write a ranked list of the top eight content opportunities to this location. For each one, include the topic, one line angle, source where you found traction, four scores, and recommended content format, like what we should create.

09:09And then log it to our, um, log files. This basically just keeps track of what the agents are doing so that we can go back and look at it, and they know, like, where each one is in the process. Cool.

09:19Now while that one is going, we are going to hit command n again and start a third agent, the growth agent. I'm going to here where I have these prompts.

09:27Again, I'll share all of these with you, so don't worry. Now this one is meant to start taking action once the other two have been completed. But the cool thing is as we loop through it continuously, um, it works in parallel with those.

09:40So you're the growth agent. The quiz has been built. Content research is done.

09:43Your job is to do everything a smart marketing hire would do in the first forty eight hours after product launch. So read these before doing anything. Then you have four tasks.

09:51First is a site link audit. So we need to figure out where should we link to the quiz on the site. Now, if we had a real website fully built out, we would link to that right now.

10:00Um, we don't because this is just a demo. I just made this up. So I'm going to say, think of a normal pickleball e commerce site.

10:06Identify every page or section where a link to the quiz would naturally fit. And then for each placement, write the exact location, the exact copy to add, and the reason it fits there. Then we're gonna write a launch email announcing the quiz to our existing customer list.

10:19So include subject line, preview text, full body copy, CTA. It's casual. It's fun.

10:24It's pickleball obsessed. We have social captions. Write three captions, one for Instagram, one for Reddit, and one for a Facebook group.

10:30Each should be native to the platform. The Reddit one should not read like marketing. The Instagram one needs a hook in the first line.

10:35Save all three to outputs slash social social captions. Next, lead magnet recommendation. So based on the top opportunities and content ideas, recommend the single best next lead magnet to build to include the title format, all of that stuff, three sentence description of what it does, why it will outperform or complement the Harry Potter quiz.

10:55Save it to here. Output slash next quiz recommendation. And then looping.

10:59After completing all four tasks, evaluate. Did the site have obvious placement opportunities that were missed in a previous cycle? Are any of the social captions similar to ones written in a prior run?

11:08Flag any repetition or diminishing returns in a file called output slash growth agent notes. This keeps future cycles fresh. When all tasks are complete, log growth agent complete.

11:18Now let's go check-in on our agents. So the Harry Potter quiz, one shot at it, and it's live.

11:25Let's check it out. The sorting paddle. Nice.

11:27Nice. Chris's Guide ten ten. Your partner looks at you.

11:30What do you do? You show up to open play, and the vibe is chaotic. Balls flying everywhere.

11:33Someone's blasting EDM. I observe for a few minutes to identify who the real players are. The kitchen game, the non volley zone for the uninitiated.

11:41Nice. Is best described as super aggressive, always. Okay.

11:45And I'm just gonna pick random questions here or random answers here just so we don't waste a ton of time. Sorting hat is ready. Enter your email to reveal your Hogwarts house playing style breakdown and your perfect paddle match.

11:55Let's go. This is awesome. And I got Gryffindor.

11:58Let's go. And the perfect paddle recommendation. So this is pretty cool.

12:01This was one shotted. I think this would be nice for a pickleball website. Obviously, it's not perfect.

12:05It's got the kind of AI look, but you guys get the point. So that's done. Perfect.

12:08Now let's go back to the second agent. This is the scout. So it found a bunch of opportunities and then ranked them.

12:17So we've got things like, is my paddle legal? An intermediate upgrade trigger guide. That's cool.

12:21Injury prevention gear guide. Shoes when the upgrade pays off. Women's gear content.

12:26Beginner starter kit quiz. Yeah. So I think this is pretty good without looking too in-depth.

12:32This seems solid. Now let's go check on our third agent. It is still working.

12:36So we'll give it a few seconds. I'm just going to speed this up. Okay.

12:39We are done now. So the post launch marketing strategy, this is from the growth agent, is finished. Here's what it came up with.

12:46Task one site edits. Let's open that here. It's came up with 12 placement recommendations and then the content for each of those.

12:54So, know, home page, uh, sticky nav bar, category page, etcetera. On the home page, place it directly beneath the main home page headline for any product grid or featured categories. Copy to add which Hogwarts house matches your pickleball personality.

13:08Cool. It didn't have the full context of the quiz, it got that wrong, but that's nice. Yeah.

13:12And then a bunch of those. Cool. Launch email.

13:14Launch email. Okay. Yep.

13:15Ernie. Yep. That's that's pretty funny.

13:18Cool. As well as copy notes. That's awesome.

13:20Social captions for Instagram, Reddit. I've been playing for about three years now, and I was bored the other night thinking about why certain player personalities are so consistent.

13:28Okay. Nice. And then next quiz recommendation, is your paddle past its prime?

13:33That's that's pretty solid. Why it will outperform supporting data points? Awesome.

13:37This is great. So now, we you've seen those three agents go through things. But the thing that I wanted to show you is if you don't wanna be the one orchestrating all of this, you wanna have an orchestrator agent, here's how we would do it.

13:51We would use a prompt like this, which is you're an orchestrator agent managing a fleet of three sub agents for a pickleball e commerce business.

14:00Job is to delegate work, monitor outputs, and synthesize results into a unified action plan. Before doing anything, check if slash outputs next steps exist. If it does, read it first.

14:09It is your memory from the previous cycle. Use it to understand it was already completed and what still needs work before delegating anything. So we've got our goal, grow the pickleball e commerce site by launching new lead magnet, researching content opportunities, and executing forty eight hour post launch marketing push.

14:25So then it's going to spawn these agents. Now, it gives a prompt here as well, which is the exact prompt that we used before. You're the builder, etcetera, etcetera.

14:34So it does that for the first one. Then it does it for the second one, the scout. Then it does it for the third one.

14:40But then where it really gets cool is it synthesizes all of this, and it writes the next steps of a summary, top three actions to take this week, and what the next loop cycle should focus on. So this is where we start looping.

14:55After we're writing the next steps, file evaluate these three conditions. Are there at least three unacted content ideas in the document?

15:03Is the site fully linked to the quiz? Is the next lead magnet defined? And if it's not met, it's going to keep doing that.

15:09From there, if we wanted, we could spin out even more agents to help with some of this stuff. So we could have a copywriter agent that is creating blog posts for all of these, and that cycles back in. And whenever we've finished writing all of the content ideas, all of the blog post ideas, then we spin out new ones.

15:25And so now you can see how this starts to build on itself. So that is one real world example, but I don't just wanna stop there because I wanna go through a few more examples to make this really concrete. And so we're gonna do a quiz.

15:36I am going to present you with four different real world scenarios that you might come across, a friend might come across, your company might come across, and you're gonna think through how would I build a team of agents to loop this. Step one, the freelancer. You're a freelance designer with six active clients.

15:50Every Friday, you spend two hours writing status updates, digging through folders, summarizing progress, personalizing each message. It's the same task, you don't like it. So how would you loop it?

15:59I'll give you three seconds to think about it, and then we're jumping down to what an answer could look like. So you have a loop that runs every Friday at 4PM automatically. It reads your project folders to see what changed, loads up a skill file that knows each client's name, project goals, and preferred tone.

16:15So this is like the memory of who your clients are, um, and how we want to write. Drafts a personalized status update, drops a finished draft in a review folder, so you look through it. Obviously, keeping the human in the loop for quality assurance makes a lot of sense, and then logs what was sent so that you never have to repeat this.

16:32And a loop condition would be, did every active client get an update this week? If yes, you're done. If not, it'll keep going.

16:38Cool. Number two, you're a student. You're studying for exams.

16:42You're trying to stay current in a fast moving field like AI. New papers drop daily, new tools every week. You feel behind.

16:47You can't read everything. So you wanna stay sharp without spending three hours every Sunday doing research. How would you loop them?

16:53Give you a second to think. And an answer is we loop every Sunday while you sleep, searches for the five biggest developments of the past week in your topic.

17:03It scores each one by relevance to what you're actually studying, filters out anything below your relevance threshold. So if it's not relevant, it won't show it to you. Writes a plain English briefing.

17:12What happened? Why it matters? And then checks the past three weeks of briefing so it doesn't repeat things.

17:16Now loop condition, are there two plus genuinely new developments this week? If not, dig deep. That's a pretty cool one.

17:22Number three, a shop owner. You sell products online. You've got SKUs.

17:25You know you should be updating descriptions, fixing what's not converting, pushing your winners harder, etcetera. But you don't do it because it takes a long time. How would you loop this one?

17:34Well, you could do it once a month. It runs on the first, reads your sales data, what sold, what didn't, identifies the three products with high traffic and low conversion, and then rewrites those product descriptions with better hooks and clearer CTAs.

17:47Writes promotional copy for your top three performers to push those harder, and then logs every change made and the reasoning behind it. And then the last one is if you are a creator. You have a YouTube channel.

17:57You post once a week. You have a bunch of video ideas, but you never know which one to make next. You pick based on vibes.

18:02Sometimes it works. Sometimes it doesn't. How would you loop it?

18:05Well, you could have a loop running every Monday morning before you start work. It reads the full ideas list, pulls the last ninety days of video performance, identifies which topics over performed and which flopped, checks what's trending, scores every idea, outputs a ranked top five, as well as flagging ideas that competitors are already recovered.

18:22Now, as much as I love AI, I don't wanna pretend like it is always perfect. So in certain scenarios, you have to have an understanding of what the AI is good at and what it's not good at.

18:34So you should try something like this. But for example, with YouTube videos, as someone who spent a lot of time creating YouTube videos, it's not always perfect at knowing what topics are really good and what's not. So you have to think for yourself, think critically, keep that in mind.

18:49But even if you have this and you can just look through it for a bunch of ideas and someone else's opinion, that's really helpful. So start building it, and then you can add in your own thought as well. So that's the idea of looping.

19:00Hopefully, those examples helped. Now you have a clearer picture. If you have a Cloud subscription or a Codec subscription, you can start testing this using it right now.

19:09It's extremely helpful and it's the future of how people are going to work with AI. If you have any comments, drop them in the comments. I'll try to respond.

19:16And I'll see you next time.

The Hook

The bait, then the rug-pull.

Boris Cherny, who built Claude Code, and Peter Steinberger, who built OpenClaw, have both said the same thing publicly: they stopped prompting. They write loops now. This video is the clearest explanation of what that means for everyone else.

Frameworks

Named ideas worth stealing.

01:28list

The 5 Stages of a Loop

GOAL (set by human once)
DISCOVERY (agents find what needs doing)
PLANNING (break into clear steps)
EXECUTION (parallel agents work simultaneously)
VERIFICATION (did this accomplish the goal?)
SHIP or ITERATE

The universal structure every agent loop follows, with external MEMORY persisting state across cycles.

Steal forAny repeatable workflow you want to hand off to an agent system

02:45model

Open vs. Closed Loop Decision Framework

Open: wide search, agent discovers, burns massive tokens, unlimited budget only
Closed: bounded goal, human builds path first, clear eval at each step, normal budget

Simple two-option framework for deciding how much freedom to give an agent before it starts costing real money.

Steal forAny agentic workflow — closed loop is the default unless you have unlimited tokens

13:38model