Everything Ryan Carson said on any show that made the record, most notable first. Each card names its show and opens the statement there.
Carson: Autonomous agent tasks must fit within a single context window
“So this is the other big unlock with working with agents is that they have a context limit, right? So you know, with Opus, you're looking at about 168,000 tokens. You have to be picking chunks of work that can be fully completed within that context window.”
Ryan Carson Built an Entire Software Feature Autonomously Using Ralph
“I actually built an entire feature with Ralph, and these are the steps I took. I created the PRD, I created the user stories, and then I started Ralph, and then I'm just going to show you an example of kind of what this looks like for real.”
Carson: Autonomous agents won't derail if constrained by atomic user stories
“People are afraid of like, what if the agent runs by itself? Is it going to go off and do crazy things? And the answer is no, because you gave it a clear user story with clear acceptance criteria, right? It's going to be actually a pretty small thread.”
Carson: AI agents should compound intelligence from mistakes via markdown memory files
“Your agent should be getting smarter every time it makes a mistake. And by updating agents.md, you are going to get that long-term benefit. This isn't just during this iteration. You're going to benefit every time from now on that you use amp or clog code.”
Carson: Ralph loop isolates context by spawning fresh agent threads
“And what's happening is you're getting a fresh loop every time. So you're getting a brand new thread or a brand new instance of cloud code every time. So you're starting fresh with a brand new context window starting from fresh.”
Carson: Autonomous AI Agents Require Atomic User Stories and Clear PRDs
“These two steps, writing a PRD and converting them to user stories, this is where you should spend a huge amount of time. Like you should spend an hour on this, right? It's very, very, very important that you get your PRD right, and that your user stories are …”
Carson claims Treehouse revenue is growing faster than ever
“We're growing faster than we ever had.”
Ryan Carson: Founder-CEOs must personally sell before hiring any sales staff
“If you haven't built the sales team yet, and you are the founder CEO, you should be doing sales. So that's the first thing you got to do. Do not try to hire people. Do not try to do SDR. Don't do any of that. Go out and sell. And look and learn the sales cycle…”
Treehouse raised ACV from $1,800 to $90,000 via founder-led sales
“This is how we raised our ACV from 1800 bucks to 90,000 dollars, right? It is by me getting in, in the dirt and figuring out what actually is valuable.”
Carson: Treehouse consumer retention is better than Pluralsight and Lynda
“You know, we have pretty good retention, you know, a little better than Pluralsight, a little better than Linda, but still it's not a rocket business.”
Carson is rejecting inbound investor interest to focus on core repeatability
“I mean, I have tons of inbound, thankfully, and, but I'm just saying, no, we just want to focus on building the core business and making it repeatable.”
Carson has no short-term plans to generate personal wealth from Treehouse
“So, I mean, that, that I have zero need or intention to generate more profit or more wealth for myself short term.”
Carson: InVision expanded Treehouse Talent Path deal tenfold within two months
“Closed it, and within two months, they expanded to a deal that was 10 X the size of the current one.”
Carson: Enterprise buyers do not care about Treehouse certificates, only skills gaps
“And what's interesting is they don't care about any treehouse certificates or points or anything. They just want to close skills gaps.”
Carson: Treehouse competitors have raised 10x more capital
“You know, our competitors have raised 10 X the amount we have.”
Carson: Standard CAC to LTV is a vanity metric without gross profit
“We think people who focus on CAC to LTV is, it can be a vanity metric period because you're not actually focusing on, on gross profit LTV. So actually how much, Profit are you driving the LTV and what is the fully loaded CAC, which means you got to include you…”
Carson claims the average developer stays at a startup for 10.8 months
“You know, the average stay for a developer at a startup is 10.8 months.”
Carson runs up to ten cloud coding agents simultaneously without collisions
“So I often have, you know, at least five cloud agents working at once, often 10, and there's no risk that the code is going to collide.”
Carson spends $60 per run for automated Devin browser tests
“And I have a it's called end-to-end sign-up test, and that runs three times a week because it is expensive. It's probably 60 bucks in tokens because it's doing a lot of, it's doing a lot of browser testing”
Carson: Untangle runs daily Claude Opus cron jobs for marketing recommendations
“Every day I have a cron job that basically looks at key marketing stats, and then it passes those stats to Opus four or five, and then it generates an email with recommendations on what is the one thing I should do to improve my marketing funnel today.”
Untangle's AI divorce agent uses state statutes to write AEO content
“So for Untangle, we have Grace. She's our AI divorce agent. She's inside of Untangle. Now, she has specific tools, right? They're very specific to our vertical, right? You know, divorce statute search, case law search, child support guideline search, right? Th…”
Carson: AMP's deep mode uses GPT-5.2 instead of Opus 4.5
“I'm actually secretly showing you a new feature. If you see deep here this is a mode. We're actually using a GPT five, two instead of Opus four or five.”
Carson: I run three AMP coding agents simultaneously
“Now, what I actually do is I actually run three of these agents at a time. So I actually have three amps open any one time, and I'm usually working on three different parts of the app at once.”
Carson: AMP will autonomously implement marketing optimizations as pull requests
“I'm thinking the next step is, is not just to get the email, but actually have AMP implement the suggestion. And then I wake up to a PR that I review where it's actually done the thing that it suggested, and then I get to review that instead.”