42:07Greg Isenberg · Interview
I got a private lesson on Claude Cowork & Claude Code
Boris Cherny, the creator of Claude Code, walks through a live Cowork demo and unpacks the 13-tip viral setup thread that got 99K bookmarks.
January 23rdThe argument in one line.
Agentic loops are only reliable when success is binary and the feedback mechanism is fixed — every other use case hands the agent your product vision and your token budget, and it will spend both badly.
Read if. Skip if.
READ IF YOU ARE…
- You build apps with AI coding tools and are wondering whether slash goal or similar loop commands will save you time.
- You are on a $20-$100/month AI subscription and curious whether full autonomy is worth switching to.
- You want a concrete, working example of an agentic loop that actually earns its place in a daily workflow.
- You have been confused by high-profile builders who say they no longer write prompts — they build loops.
SKIP IF…
- You have unlimited token budgets and are doing research-grade experimentation — this conversation is not for your cost structure.
- You already run sophisticated multi-agent harnesses with test suites and browser-use feedback; you are ahead of this discussion.
The full version, fast.
The viral framing that you should design loops instead of writing prompts makes sense for researchers with unlimited token budgets, but not for builders on $20-$200/month plans. A fully autonomous loop removes the human from the decisions that most determine whether a product is right — leaving the agent to fill gaps with assumptions that drift from your vision and drain your account. The only place loops reliably pay off today is constrained, binary-feedback work: code review, SEO pages, any task where a score or a pass/fail gives the agent something objective to chase. Human-in-the-loop is still the best loop.
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 →Who's talking.
00:00hostGreg Isenberg
00:36guestRoss Mike (Ras Mic)
Where the time goes.
00:00 – 01:23
01 · Intro
Greg frames the episode: clear definition, honest hot take, real working example.
01:23 – 07:59
02 · What is a Loop
Whiteboard walkthrough of human-in-the-loop vs. the Boris/Peter fully autonomous agent loop, using stick-figure diagrams.
07:59 – 11:32
03 · /goal Explained
All slash loop commands are the same pattern. Two problems: heavy token burn and the impossibility of spec-ing your full product vision in one document.
11:32 – 12:42
04 · The Slop Machine
Greg coins the phrase. Mic imagines the sophisticated Boris/Peter harness (test suites + browser use), then names Peter's $1.3M/month token burn.
12:42 – 18:19
05 · Code Review as a Loop
The one loop Mic runs daily: Cursor to GitHub to Greptile score to greploop skill to iterate until 5/5 or 5 turns. Only ships to production at >4/5.
18:19 – 21:02
06 · Honest Take for Builders
Loops work where output is binary — code review, SEO pages. They fail for app-building that requires mid-course real-user feedback.
21:02 – 22:10
07 · The Future of Loops
Both agree autonomous loops are coming. As of June 9, 2026, human-in-the-loop is still the best loop.
22:10 – 22:32
08 · Closing Thoughts
Wrap, credits, and mutual appreciation.
Lines worth screenshotting.
- A loop fires once from a human, then the agent generates output, reviews it, feeds it back, and continues building — no human steering after the trigger.
- Human-in-the-loop is not a limitation; it is the layer that supplies product judgment the agent does not have.
- Your spec.md covers 80% of what you want; the other 20% is product taste, and the agent will guess wrong on every edge in that gap.
- Boris and Peter operate with unlimited token budgets — their advice to build loops is rational for their cost structure, not yours.
- Peter burned $1.3 million worth of tokens in a single month; that is the honest data behind the viral loop evangelism.
- Slash goal, slash loop, and equivalent commands in every coding tool are the same pattern: fire once, do not stop, burn tokens until done.
- If you are on the $20 or $100/month plan, agentic loops should not be a consideration yet.
- AI can replicate sauce. It cannot create sauce. Keeping humans in the loop is how product judgment stays in the system.
- The slot machine metaphor: loops produce random-quality outputs and you cannot see the result until the spin finishes.
- Code review works as a loop because the success condition is explicit: get the Greptile score above 4 out of 5.
- Even a well-designed code-review loop breaks past 1,000 lines of code; the agent cannot fully contextualize a large diff.
- Loops are fine for SEO pages where every output follows an identical format — but generating 300 identical pages is not building a startup.
- The missing piece in every wide-open autonomous loop: no real user ever sees the app midway to tell you whether you are building the right thing.
- Full self-driving from Miami to Charleston with no stops — that is the honest metaphor for what a wide-open agentic loop actually is.
- Human in the loop is the best loop — the guest's closing thesis, stated clean at 21:52.
When to hand the wheel to an agent loop.
WHAT TO LEARN
The right question is not whether loops are powerful — it is whether your task has a fixed, machine-readable success condition, because without one the agent is just spending your money on guesses.
02What is a Loop
- A loop is not magic — it is a workflow where the agent checks its own output and feeds it back as input, skipping the human review step.
- Human-in-the-loop is the baseline that keeps product judgment in the system; removing it is a deliberate tradeoff, not an upgrade.
03/goal Explained
- Every spec document you write is incomplete. Product decisions evolve, and some only become clear when a real person uses the thing.
- Slash goal is a fire-and-forget command that trades your oversight for the agent's assumptions. Use it when the assumptions do not matter, not when they do.
04The Slop Machine
- The $1.3M/month data point is not an indictment of loops — it is a reminder that loop evangelists are operating in a different cost structure. Translate their advice to your budget before acting on it.
05Code Review as a Loop
- A loop is only as reliable as its feedback signal. Code review works because a score is a score — pass or fail is unambiguous.
- The 1,000-line rule is a practical ceiling even for well-designed loops. Split large diffs into smaller PRs and loop on each one separately.
06Honest Take for Builders
- The missing ingredient in every autonomous build loop is user feedback midway. No amount of planning in a spec.md replaces someone actually touching the product and telling you what is wrong.
- Binary output tasks — SEO pages, code review, format-identical generation — are the valid home for loops today. Anything requiring creative judgment is not.
Terms worth knowing.
- Agentic loop
- A workflow where a human fires a single instruction, then an AI agent executes, evaluates its own output, feeds the result back as input, and repeats — with no human step between iterations.
- Human-in-the-loop
- A workflow where the human reviews and approves each AI-generated result before the next step begins, maintaining control over direction and decisions throughout.
- Slash goal / slash loop
- Autonomous loop commands in AI coding tools that instruct the agent to complete an entire task from a prompt or markdown file without pausing for human review.
- spec.md / PRD.md
- A markdown file containing the full product or feature specification used to seed an autonomous agent loop — intended to give the agent everything it needs to build without follow-up questions.
- Greptile
- An AI code-review agent that integrates with GitHub, automatically reviews pull requests, and returns a quality score out of five alongside specific feedback on bugs and edge cases.
- greploop
- A custom Cursor skill that reads a Greptile code-review score, instructs the agent to fix the flagged issues, pushes the changes to GitHub, and loops until the score reaches 5/5 or five turns elapse.
- Meta harness
- A sophisticated wrapping system around an AI agent loop that adds structured feedback mechanisms such as test suites, browser-use screenshots, and scoring — making the loop more reliable than a naive fire-and-forget command.
- Token burn
- The cumulative API cost incurred when an AI agent runs autonomously without human checkpoints, often orders of magnitude higher than directed prompt-by-prompt usage.
Things they pointed at.
13:33toolMacroscope
08:03toolSlash goal / slash loop commands
Lines you could clip.
11:05
“AI can replicate sauce. It can't create sauce.”
00:41
“You're gonna understand why it is a terrible mistake, and unless you have money to burn, you are not to do it.”
21:52
“Human in the loop is the best loop.”
12:39
“In one month he burns 1,300,000 dollars worth of tokens.”
15:54
“This is a very closed off, very goal-oriented loop.”
Where the conversation goes.
00:00 – 01:23sparseIntro and promise
01:23 – 07:59denseWhat is a loop — diagrams
07:59 – 11:32denseSlash goal explained
11:32 – 12:42steadyThe slop machine — Boris and Peter context
12:42 – 18:19denseCode review loop walkthrough
18:19 – 21:02denseHonest take for startup builders
21:02 – 22:10steadyThe future of loops
22:10 – 22:32sparseClosing thoughts
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.
metaphoranalogystory
00:00Everyone is talking about AgenTic loops, but the reality is most people don't know what it is or how to use them. In this app, I brought on professor Ross Mike to clearly explain what it is, is it hype, is it real, and how to use it.
00:16And if you stick around to the end of the episode, he shows me the most concrete use case of Agentic Loops that you can use starting today.
00:25Enjoy the episode.
00:35Ross, Mike, welcome to the pod. By the end of this episode, what are people gonna learn?
00:41You're gonna understand what a loop is. You're gonna understand why people are fanning out about it, and you're gonna understand why it is a terrible mistake. And unless you have money to burn, that you are not to do it.
00:51I'm also going to play the other side, and I'm gonna show you a loop that I use. But the general consensus, I think, is wrong, and we're gonna talk about it. Okay.
01:00So by the end of the episode, people are gonna understand what an agentic loop is, why
01:05the most well known people in the AI industry are obsessed about it. You're gonna keep it real with what we need to know about it and what we can avoid, and you're gonna show a real use case, a real example of how to actually use Nagentic Loop.
01:19Exactly.
01:20Exactly. Alright, bro. Alright.
01:22Let's get into it. So as always, um, a lot of people love the diagrams, so we're just gonna start with diagrams. I paid, um, like like, I think $3 to get these stick figures, so I I hope people appreciate them.
01:36This is me and you. Right? This is your average Josh Moe who does not work at Anthropic or API or OpenAI.
01:44And this is Boris and Peter and anyone else who has unlimited access to models.
01:52Now, the way me and you have been working, um, this is what is called a human in the loop is you and I will prompt our, you know, computer. Right?
02:03Let's say this is our computer or better yet, I'll say this is our AI agent. Right? Whether you're using cursor, Claude, codex, doesn't matter.
02:10You are prompting it yourself. Right? You're telling it, hey, build me this landing page.
02:16You know, build this feature x y and z. You are communicating with an AI agent, a platform of your choice via a prompt. And then a result is generated.
02:26Right? A result is generated and usually what you and I will do is we will view this result, we will test this result, and we will keep on iterating.
02:37This is the loop where it goes back to us. Right? So let's say I'm working on an app and this app is a to do list app, Greg.
02:47Um, the first thing that I'll probably wanna do is I wanna build up the landing page because I wanna get this out to the public. So maybe they can sign up and join the wait list. So I'll prompt and build me a landing page and let's say I like the landing page.
02:58Next, I'll work on authentication and then once I'm happy with authentication, then I'll work on with the back end. So this is what we are used to and this to be, uh, sharp with it is called human in the loop.
03:13Meaning, it is the agent that's building, but it is you that is directing, governing and allowing things to happen. What everyone has been talking about particularly Boris and Peter, uh, they said they don't write prompts, they generate they build loops.
03:29And essentially what they're talking about is they're building a system and I'm just gonna show you here where this is the AI agent. Right? And then this is the result.
03:44But instead of a human being in the loop, the human is in the loop one time, meaning it fires off, um, said loop. But then the rest of the time, it's the agent checking, uh, it's the agent generating a result.
03:58That result is then fed back into the agent. The agent then looks at the result and continues to work. Now, this in theory sounds cool because what essentially we're saying is, hey, I'm just gonna have some sort of spec dot m d file or some PRD dot m d or whatever dot m d file, and this is gonna be like a to do list, a task list, and this is going to give all of the information the AI agent needs to build this.
04:25Now, this sounds cool and this low key might be the future, but here is where it goes terribly wrong.
04:33First and foremost, I wanna get paint this analogy. Let's say
04:38Quick break to talk about today's sponsor, CodeRabbit. We talk about Adjenta gloops in this episode, and most of them are still hype. CodeRabbit is one of the few that actually works.
04:50My team literally uses it every day. You open a poll request, and CodeRabbit reviews it automatically. It catches bugs.
04:58It suggests fixes, and it's one click to commit. It learns your team's standards over time, so it just gets better and better.
05:06They just shipped CodeRabbit review, which takes a big messy pull request and reorganizes it so you can actually understand what's changed and why in minutes instead of hour, something I've wanted for years.
05:20My team at LCA and Idea Browser love it. There's a fourteen day free trial. If you wanna learn more about CodeRabbit, go to the link in the description.
05:29We we're we're building a startup. Right? You and me, Greg, we're building a startup.
05:32We hire a very smart developer. And we tell this developer, this is the app we wanna build.
05:37These are the things that it needs. And the developer goes on and builds the entire thing without consulting us. In building that entire thing, that developer is going to have to make assumptions.
05:47Right? Assumptions of how the product looks, how it's going to feel, certain architectural decisions. There's a lot of assumptions that are gonna be made in the nitty gritty.
05:55Now you might think your plan document covers everything, but truth to the matter, it never does. There's always an edge case. There's always something that's missed.
06:04So what the developer is going to do is that developer is gonna make a lot of assumptions. Those assumptions might not be aligned with our product vision. Now you have a developer who's come with a finished product, and now there's a bunch of things in order, but it's not the way we want it.
06:18In the same way, when you have this stacked PRD dot MD file or this whatever markdown file you have and you give it to agent and you run it in this loop, meaning it takes the feedback it takes the result, takes it as feedback, and continues to generate, uh, code. What happens is you now have an agent that's going to make assumptions.
06:38And believe me, when you give the agent the floor to give assumptions, most of the time, it's going to get it wrong. But not only is it going to get it wrong, it's going to burn a lot of money.
06:49Now I say this with all love, but Boris and Peter come from a place where they have no token budgets. Right? They can burn unlimited tokens.
06:58If I had unlimited tokens, I'd be doing the same thing too. But this is not productive. This is great for research and I'll actually share a loop that I use.
07:09But this idea of construct of constructing like a meta harness where you give the agent feedback automatically, like, it gets, like, the information, the result it's generated, and it loops on it, um, it is a catastrophe.
07:23And we've tried this. Right? We had Ralph Loops.
07:25We had Ralph Wiggum. Um, there's even, like, slash goal, uh, which has been pretty popular the last couple of weeks. These are great to build prototypes.
07:33These are great to experiment with. Like, let's say you wanted to experiment with something. You wanted, like, some miniscule tool built out, but you didn't care about the nitty gritty details.
07:41These are great. But if you're if you care about the details and you don't have tokens to burn, this is the worst thing to be trending right now in my humble opinion.
07:54Uh, I'll stop right here just to make sure, Craig, I'm making sense because, uh, as you could see, I'm pretty passionate about this.
08:00So slash goal is also trending at the moment, you know. Is slash goal a loop?
08:06Like, should people think about goal
08:07and a loop? They're all the same thing. They have different names like slash goal.
08:11I know on cursor, think it's slash loop and then on another tool, it's slash whatever. They're all the same thing and basically how all of them work on a high level is you you, you know, you type in slash goal and then you give it some prompt. Right?
08:25Like, you give it some prompt and then you can also, like, attach some, you know, markdown file and you tell it, like, yeah, build this entire thing out. Don't stop until you're done.
08:36Don't make any mistakes. Again, these are cool, but the two issues are, number one, they burn a lot of tokens.
08:43Right? Um, and if you are not like, this shouldn't even be a thought if you're not on the $200 a month plan. Like, not a thought.
08:51Like, if you're on the $20 or I think there's a $100 a month plan, you shouldn't even think about this. Right? Because it's just gonna burn your token usage.
08:59Um, number two, you think your plan is good, but it's not. Because it's impossible for you as a human to contextualize every single detail about the product that you want, um, in one document.
09:12Right? Things evolve, trends change. Um, you know, one day liquid glass is cool.
09:17The next day, we're changing how liquid we want it. Like, it's very impossible for you to fill, like, uh, your thoughts and exactly how you want the product to be one in one in one document.
09:29If anyone works in service, whether you you run an agency or you like, for example, we develop software for other for other people and companies. It I I we try all the time to get all the thoughts out of someone's head. There's always something.
09:42There's always, oh, you missed this or I wanted it like this. This is what I meant. How much more do you think an AI agent's gonna understand you if we as humans have hard times understanding each other.
09:51Right? So this should only be used in the following.
09:57Experimentation, like, me do a new line. Experimentation.
10:01Right? Let's say you wanted to, like, I'll I'll with you a fun cool little tool I built the other day, Greg. I was doing a talk and I wanted to build an Among Us simulator for for AI models.
10:12Right? Basically, it's this game where there's one bad guy, one impostor, and everyone's trying to figure out who it is. And I wanted to have like my own benchmark to find out which models like are good at lying.
10:23And I didn't want it I didn't care about the details. I didn't care about how it looked. I just wanted the simulation and the benchmark to work.
10:29So I told it, I want this simulation. I want this benchmark. I wanted to do this.
10:34Go and do it. It took about, I think, an hour and a half and it got it done. Now there were a lot of details that I had in mind that it completely got wrong, that I didn't specify in the initial instruction set.
10:47But guess what? Because I didn't care about the what it built in a sense, like, I didn't really care for the details. It was a great thing.
10:55I didn't spend a lot of time. I just got slash goal to take care of everything. But when you and I are trying to use AI to build something meaningful, I 100100100% stand in the fact that the human still needs to be in the loop.
11:09. So if I just have these giant loops running, um, and then once they're done, maybe I'll go in and fix things up.
11:20Sure. You can make that argument, but I hope you have money to burn. Right?
11:24Like and that's the ultimate thing. This will burn money.
11:29It sounds cool, but it'll burn money.
11:32What I'm hearing you say is that the loops are going to create a slot machine.
11:42That that's basically what it is. Now, I have no doubt the Boris' and the Peters' are building very sophisticated like loops.
11:49Like, I can imagine, like, let me drag this over to here. I can almost imagine they have something like again, I I I I'm not sure. This is just me guessing, but I can imagine they almost have like some sort of, um, test suite, right, where like they write, uh, tests for the agent to run the code against, so it's a certain type of quality.
12:09I'm sure they also have some sort of browser, uh, browser use capability, so the agent can see the page live and can take screenshots.
12:18Like, I'm sure they have a insane harness or meta harness around the agent so that this loop can be more successful than the average loop.
12:27But at the end of the day, the one argument I'll fight back with is this is going to burn a lot of tokens. And if you don't believe me, all you have to do is look at Peter's tweet where in one month he burns 1,300,000, uh, dollars worth of tokens.
12:42But I don't wanna sound like a, you know, like a I don't know what the word is, like a doomer, like, oh, like like an old guy, like, decent suck. There are use cases and I'll share one, Greg, if that's okay, where my, uh, my code review process is a loop, and I'll explain how it works.
13:02So, um, I use cursor for the most part, not sponsored. I use cursor for the most part as my, uh, harness of choice. And with cursor, I will use GitHub as my source control.
13:17Basically, a place where I store code, version code and all that stuff. And every time I push a feature, like every time I build a feature, I push a feature or whatever the case may be, I am pushing code to GitHub. And in GitHub, I have a code review agent installed.
13:36There's many kinds. My particular one that I use is Greptile, but I know people use CodeRabbit, Macroscope. They're all great.
13:42I use Greptile. Um, and what happens is whenever I push a feature to GitHub, the code that's being pushed to GitHub is AI generated.
13:52But then I have a code review agent that reviews the AI generated code. And what's cool about Reptile is it gives me this review. Right?
14:00It'll be like, oh, you missed this. There's this security thing. You this is broken in this edge case.
14:06It's pretty it's pretty good. But my favorite thing is it gives you a score, a score out of five.
14:11Right? It could be two out of five, one out of five, five out of five, four out of five, whatever. It's a score out of five.
14:17And what I the the the mental model I now have is I will not push anything to production, meaning I will not allow code to go live unless the score is greater than four out of five. Right? If the score is not greater than four out of five, this code needs to be reviewed.
14:35Now here is where I loop. I have this skill called grep loop. Right?
14:41And basically, it again, I I don't want people to think it's complicated. I just want you to understand where loops make sense. It's basically a skill that tells the agent, oh, check GitHub, read the review, and then fix the review, and then push to GitHub.
14:54So what happens is when I see a score, let's say, I got a two out of a three out of five, again, my rules, um, is that it has to be at least four out of five and greater. So what I'm gonna do is I'm gonna go back to cursor, I'm gonna write greploop.
15:09And then when I write greploop, what happens is cursor reads the review that greptile wrote on GitHub and then it feeds the review back into Cursor. Cursor then makes the changes, pushes the changes to GitHub, and then waits for Greptile to do a new review.
15:24Every time you push to GitHub, Greptile does a new review. If the review still is a three out of five, guess what happens? The loop continues, and then more changes are made.
15:34And then let's say it's a four out of five. It doesn't give up. It keeps it takes the feedback and then pushes it back to GitHub.
15:41It won't stop unless it's taken five turns and then it'll give up, or it won't stop until it gets a five out of five. Now, this is basically a loop. But if you notice this, Greg, this is a very closed off, very goal oriented loop.
15:57Essentially, I have a feedback engine. Right?
16:00I have a a code review agent that's giving a score. What I'm telling cursor is read the review, understand it, and get that score to a five out of five. This makes sense for code review because there's a fixed feedback loop.
16:15But when I'm building an app, again, I have no idea what I want completely in that very moment. So it's very hard for me to generate a loop on an app that I have in my mind, but I I can't even fully visualize just yet.
16:29Now if you're great at visualizing, you're a master, you never miss details, you've never forgotten your auntie's birthday, you don't forget your wedding app, you're just perfect, and you have a million dollars, go ahead and build loops.
16:42Mhmm. But for me and myself, the only place a loop makes sense is in a very confined constraint process with a very fixed feedback loop, a very defined feedback loop, and that's in code review.
16:56And can I be honest with you? This loop actually quite breaks at times.
17:01It's not perfect. And you know when it breaks, anytime I push over 1,000 lines of code, one k lines of code, like, if the if the code that it has to review is more than 1,000 lines, I can almost never get a five out of five because it's too much code for the agent to fully review and contextualize and understand.
17:20So even in this fixed sort of ecosystem loop that I have here, even here, there's there's there's reasons and places for it to break.
17:30Right? So every time I push a change, I have to make sure it's one k lines of code or less or I have to tell the Asian cursor split this into multiple p r's, multiple code pushes so greptile can review.
17:44I say that all to say, I'm not a hater, but loops just don't make sense right now, especially for building apps.
17:52They make sense for code review. They maybe make sense for, like, you are trying to do some SEO and you have, an SEO formula and you want, 300 pages generated and all the pages look and sound the same. Go ahead.
18:06But for anything that requires a slight bit of creativity, unless you're looking to donate money to companies that are about to go public at trillion dollar valuation, this just just this just doesn't make sense to me.
18:19The person listening to this podcast, I mean, it's literally called the Startup Ideas Podcast, they're building apps. You know?
18:26So what they're doing, you know, is they wanna create an app, a website, a startup, a SaaS, a micro SaaS, an agent for a startup that has the highest likelihood of success. And in order to do that, you have to show your app to people in order to get feedback.
18:45Mhmm. So what's missing from the the loop is there's no sharing your app for feedback halfway through.
18:53Right? You're just pressing or slash goal or you're just like basically, it's think of it as like full self driving.
19:01You're going from Miami all the way to Charleston, South Carolina, and you're pressing go. And there's no, like, you're gonna go off and you see a really, you know, cute diner on the side of the road, and you're gonna go order a fried chicken sandwich.
19:15You you can't. You're you're on this, you know, you're on this ride, and whether you like it, the train has left the station.
19:23That's what this is. And so
19:26my my belief on loops is actually a lot similar to yours. By the way, I loved your rant. You know?
19:32I apologize. And again, if any of the companies are planning on sponsoring whatever, I love you guys. But this is just I can't lie to the people.
19:39You know what I mean? Like, I gotta be honest. Like, I've seen a lot of people excited about this.
19:43And an idea this is cool. But, like, I know my some people got $20 subscriptions, $100 subscriptions. This will burn through that, and it's not productive at all.
19:53So, yeah, just have to keep honest. I think where where the output is binary,
19:58meaning black or white with no creativity, there is a room for loops. That's my I so that was my when I was reading all that was going on, like, was like, okay.
20:08To your point, like, with CodeRabbit or Greptile, the you know, CodeReview or SEO or know, those pages, like, it's it's binary.
20:18Either they did the job or they or they didn't. So I think there's room for it. But for for the people listening to this that are like, I'm gonna go build a startup, and I'm gonna I I need a loop to go build that startup.
20:33You know, make me a million dollars. Make no mistakes. Cross.
20:36Right? Like, that's where I think that this is a little bit misleading. That being said, to the credit of Boris and to the credit of Peter and to the credit of the people who are talking about it, I do believe that we will get to a point 100%.
20:54That Some point in the future. 100%. Future.
20:57Right? That this will be possible.
21:02Just not now. Just Alright. Maybe not as of recording June 9 06/09/2026.
21:07And, again, like, I don't fault them. Like, I don't think they're malicious at all, but it's like, if you have, like I'm telling you, if I had a limited token budget as well, Greg, why the heck would I prompt?
21:19Like, tokens don't matter to me. Right? And it makes sense.
21:22They have to experiment. They have to like, they they they need to work on self healing agents and all that type of stuff. So their position makes sense.
21:30My issue is everybody else who's, you know, creating content and teaching these things saying, oh, this is creme de la creme. It's like, no. Unless you want to donate to trillion dollar companies, this is not it right now.
21:43It could be it in a month and I could look like a fool, but right now, this just does not make sense. Human in the loop is the best loop. Thanks for clarifying
21:52it and and keeping it real with us. Ross Mike is his YouTube name. I'll include a link where you can follow him in the show notes and in the description and on X.
22:05Thank you for coming on here. You're you you were just the person I needed to come on the show to just say, go off, king.
22:12And and and, like, you were you were a loop you were a loop in the sense that you were the looper and just explain this clearly and keep it real.
22:22Just don't stop. Just wanna Don't stop.
22:25Well, I appreciate you for having me, Greg, as always. It's a pleasure, man. Thank you.
22:29I'll I'll see you next time. Bye, everybody.
The Hook
The bait, then the rug-pull.
The tweet that started it all said you should stop writing prompts and start designing loops that prompt your agents. Ross Mike sat down to explain what that actually means, why it breaks most budgets, and the one daily exception that earns its keep.
How they asked for the click.
MENTIONED ON CAMERA
FROM THE DESCRIPTION
PRIMARY CTAWhere the creator wants you to go next.
