Adam Sandler

I'm honored to be joining forces with @IMJustinBrooke to lead a 2 day acceletor-style deep dive on building intelligent marketing systems and agents. The curriculum is designed for people who are ready to roll up their sleeves and go, those who feel the urgency of learning platforms like Claude Code or Openclaw. The event is April 24th and 25th in Nashville TN, please DM Justin to grab a seat, spots are limited to 18 attendees. By April 26th, you'll be developing your first Agent employees.

@IMJustinBrooke I like unfurl

@TheViableEdge Yea install/setup/deploy not sure what the right word is for these. Feels more like deploy.

Have you heard about Hermes agent? What about Paperclip? I’ll be showing my setups for these two new AI tools in Nashville. Let’s start with the basics… Hermes is the evolution of Openclaw. Works just like 🦞 but with even better self learning, persistent memory, and wayyyy less security issues. Openclaw = Charmander Hermes = Charizard You can install them on a machine if you like, but I still recommend and personally use a $24/mo Digital Ocean droplet to host mine. Here’s where Hermes outshines its predecessor… If you teach Hermes how to write like you, after 10-12 articles or social posts it’ll be scary good. Its self learning ability is far above anything else out right now. You just have to give it some feedback in the beginning and it keeps everything saved. It even builds its own skills to achieve the task better and better. Perfect agent for managing email support because it self learns and with persistent memory it actually remembers each customer. Good luck asking your VA to do that. Or let it manage your community and it’ll start answering questions better than you and remember each members projects. But then there is Paperclip. This one’s different. It’s an orchestrator, not an agent. Think of Hermes as a star employee and Paperclip as the CEO. Paperclip assigns work to Hermes (or any model you plug into it). You give Paperclip a company name and mission statement. Then YOU act like the board of directors. Paperclip develops an org chart and the CEO agent asks you permission for things like budget and plans it comes up with. Mine acts as a marketing agency. All my brands and different products are the clients of this marketing agency. It’s building SEO skills, social media skills, and Google ads skills. All runs autonomously! It just sends me text messages for me to approve or deny. But here’s my next step… I’m going to try to teach Hermes how to replace me as the “board of directors.” It should work, since Paperclip has an API. Instead of asking me to approve or deny, it’ll ask Hermes. Which after I train it for a few rounds, the self learning aspect will kick in. It’ll learn how to think like me and choose approvals and denials that I would have chosen. How are you staying on top of all of this stuff? Are you? Join me and @TheViableEdge edge in Nashville for a 2 day event. I’ll show you how to install these things the super easy way. Because they have a CLI!! If you don’t know what that means, then you REALLY need to come out to Nashville.

@IMJustinBrooke This is a really clean and clear explanation, thanks for breaking it all down. Definitely interested in the Build Notes!

Before they change the definition on us… Here’s what ANI vs AGI vs ASI means (or meant) originally. ANI is what we are currently experiencing, we just call it AI though. It stands for artificial narrow intelligence and means it’s AI that’s really good at specific tasks. Kind of like how Claude Code is really really REALLY good at coding. Or Claude Sonnet being great at writing. This is why there are soooooo many AI tools. Each has to be specifically trained (fine tuned) on a specific task. Then there is AGI. This is the one all the big tech companies are racing towards. It’s the reason HUNDREDS of billions of dollars have been invested. AGI stands for artificial general intelligence. Meaning AI will be better than humans at most/all “general” tasks. This is the “moving” definition. I say moving because 4yrs ago when I first got into AI, it meant all general tasks. Today it’s been bent a little to mean “most” tasks. And I believe in a few months Anthropic (maybe another company) is going to claim AGI but the definition will be bent even more. My needle points to Anthropic specifically because they invented “agent skills.” Millions of people are uploading their S.O.P.s, best practices, proprietary methods, and recipes into Claude via their skills. Now they have plugins which are like advanced skills that do a wide variety of things. Soon (next two big updates) they’ll probably have categories or roles. It’ll be an update that says their AI can do all the skills in whole roles or categories. Like marketing, or finance, or health etc. We are super close to this. It could be a month away. It could be tomorrow. But I’m thinking Fall season is going to be massive. These companies will want to have record numbers for the year. They’ll work hard all summer for Fall launches to give enough room for news cycles and 2mos of growth. Which leaves us with ASI. This is artificial super intelligence or what y’all would refer to as SkyNet. That’s the one that will be better than humans at pretty much everything, but not marginally better… exponentially better. So much better that our minds can’t even grasp at how much better because we don’t yet know what we don’t know. For example, some people can taste colors. It’s a real disorder or phenomenon (forgive my ignorance, not a doctor). What does yellow taste like? What does light blue taste like? Your mind has no context for it, so you can’t even imagine. ASI will literally be off the charts. It’ll be like a copywriter who can predict in real time the exact words to use for every individual visitor to your website because it just read every social media post they’ve ever made. (Someone build that!) This will require quantum computing. The tech of today is just not physically possible to get there. Really we can’t get to AGI either, but the definition is covered in more baby oil than P. Diddy has in his basement. It’s slippery and getting more slippery by the week. I won’t be surprised if someone from big tech tries to beat everyone with a blow out summer announcement. But just remember ChatGPT was announced in November. So was Kimi K2, Gemini 3, and Apple AI. It’s not accidental. Lastly, one more prediction. Next year will be the year of local LLMs. The models are good enough now to compete with the big subscriptions. They’ve also become efficient enough to run on common machines as well as computer companies developing AI ready consumer computers. The big 3 will raise prices, it’ll bring local models mainstream instead of just extremists like Finn and his cult following. They have to raise prices, it’s getting harder to scale on hardware alone. I’m hoping for a $1,000/mo mega max plan. I’ll be first in line. Already tapping out my $200/mo plans on Manus, Gemini, and Claude. Launching “Build Notes” this week. It’ll have news like this plus detailed AI recipes. Interested? Let me know…

Fun with Claude Code in desktop mode

@bradsmith @IMJustinBrooke @alejandroreyes @pena_austin @BishopBarron I'm just happy to be here, the mutual learning and support is like oxygen.

@IMJustinBrooke @TheViableEdge @alejandroreyes @pena_austin @BishopBarron Yea we could build even more magic and help so many people

My dream team… @TheViableEdge @alejandroreyes @bradsmith @pena_austin @BishopBarron Real ones. 💯

Dregwick, the Uncommon Cactus. Okay, you've got my attention.

@fin465 Guide

Now that we’re done at YCombinator, we’re revealing how we went from 0 → $10k MRR in our first 30 days, using only ONE channel (step by step). We spent less than $100 and didn’t have any paid ads, SEO, waitlist, or content marketing. Instead, we sent 50-75 highly targeted cold emails a day. Cold email is the most underrated channel because it's hard to get right, but if you figure it out you can sell ANY B2B product. Here's what we did from start to finish: STEP 1: Build an ultra‑specific customer profile at both company and person level. If you do this right, you can mess everything else up and still succeed. The goal here is to create such a perfect customer, that if they heard about your solution they would have no choice but to say "tell me more". Step 2: Build your list After you create this customer profile, find the companies that meet this criteria. Find 30–50 target companies on LinkedIn, then grab decision‑maker emails via Apollo/Wiza. STEP 3: Writing a killer email I used to run an outbound email agency and we'd send 50k+ emails/month to book b2b sales calls via cold email. Here are the basic principles of cold email writing that I always use: -Keep it 5-8 sentences. 70%+ of emails are read on mobile, so make sure they get most of it from that screen view. - Never write more than 2 sentences without breaking up the lines. People skim, and that’s the best way to keep their attention - DO NOT talk about your product’s features. - Instead, talk about the person, their company, and their pain points. STEP 4: The call I took 493 sales calls in Origami’s first 3 months. Here's what I learned: The 2 biggest goals for this call are - Figuring out the customer’s problems - Getting the customer excited about your solution Unless you already have PMF, it doesn't matter if you have a full built product. You still need to spend 90%+ of your time figuring out what the customer actually needs. In the early stages, you can even offer a full refund if they aren’t satisfied to give them maximum confidence and get your first few deals over the line. STEP 5: Closing/After Congrats! You cracked cold email. This was the exact approach we used at Origami to get our first $10k MRR, and the highest converting outbound approach I’ve seen when I ran my agency. I posted the stats in my prior tweets, but in our first 40 days we sent 3119 emails (~77 per day) and got a 5.3% response rate, resulting in demos with 64 founders at companies within our ICP. This resulted in ~$22k new MRR by the time our sales for all of these calls had closed. The best part is that once you nail this process, you can automate it. We've got our Origami AI Agents (@origamichat) finding new customers 24/7, which frees us up to explore new channels and focus on scaling. CONCLUSION This is a very short version of my guide. The full guide I posted on X last year (@fin465) hit 800k impressions and 10k+ bookmarks. If you want me to DM it you, comment GUIDE.

Full 30-min build walkthrough here: youtu.be/SaOTZWJBGtU?si…

I asked Claude Code to build me a full marketing operating system with AI agents. It asked 6 questions, then built 27 files, 8 agents, and 5 workflows. This is what agentic AI looks like when you stop talking about it and start building.

What are your Claude Code and Cowork Computer-use use cases? I used Dispatch to mute my laptop from across the room, that was cool but I would love to hear about productive use cases and ideas on how to leverage this capability.

@AdolfoUsier @opencrabs I think the 72 hour limit is probably a great safety net for many of us who are learning, and still very useful for project based needs.

@TheViableEdge 72 hour session limit is the catch nobody mentions tho. ambient agents that survive session death are the move - I built @opencrabs with heartbeat cron so agents wake up do the work sleep. different mental model than /loop in a CC session

Here's a simple Claude Code automation tip you can use right now. Claude Code can set up cron jobs, think of it like a scheduled task that will trigger based on criteria you set. This cron job will be persistent as long as the session remains open, for up to 72 hours (a guardrail built in to Claude Code) There are two ways to do set up these tasks: 1. using the /loop command, example: /loop "prompt" 10m (10m = 10 minutes, or whatever interval you want). 2. Prompt via natural language. Example prompt: "Set up a cron job to poll [directory] every 10 minutes. When a new file appears, use the [skill] to update the website with new data" A simple use case: a Cron job can be set to poll for changes in a file or directory, and then set to perform some action when a defined change is detected. Here's a use case: I had a CC session running and reporting agent evals, a process that can take 30 - 60 minutes per run. When the run finishes, it creates a new report that I publish on my website. I created a separate Claude Code session and created a cron job to poll the report directory every 10 minutes to detect when a new report appears. When the report finally appears, the cron job is set to trigger a skill that updates my website with the new data, and then slack me the results and confirmation that the website changes were pushed.

Absolutely, and compounding results means you need a reliable, measurable feedback loop, quality control, consistent and ongoing evals to prevent drift with changing context. Not enough people talk about these things, and it's something we'll be unpacking throughout the Nashville event.

@IMJustinBrooke @memecoinslunar @TheViableEdge AI agents are powerful. The real advantage is how well you orchestrate them into a system that compounds results.

This is the mansion I rented in Nashville for April 24th and 25th. I have 18 seats, here's what's happening... The gap between people using AI at a 2024 level (ChatGPT prompts, custom GPTs, copy-paste into a Google Doc) and people using AI at a 2026 level (command line agents, context engineering, MCP integrations, orchestration) is enormous. And it's getting wider every month. So me and my buddy Adam @theviableedge rented a mansion in Nashville. April 24-25. Two days. 18 founders. This is not a seminar. You won't sit and take notes. Every session ends with something new running on your laptop that wasn't there when you sat down. You'll leave with a Prediction Agent, a Ghostwriting Agent, 85 pre-built marketing skill files, MCP connections to your actual tools, and the knowledge to build new agents on your own. I only have enough seating for 18 people. If you want to be in one of those seats comment "I'm in" and my assistant will DM you the invitation link...

@francois352 Thank you sir!

@TheViableEdge Great tutorial, easy to follow and implement, as usual

Here's a beginner tutorial on how to build a marketing agent system With Claude Code Full Walkthrough, written steps and video: viableedge.com/yt/marketing-a…

@MicalJohnson Cool, I'll DM you a link shortly, I'm away from the computer for a bit.

@TheViableEdge This would be cool. I’m game to test it out for you.

Created a one copy and past prompt for Cowork that sets it up as an assistant, which does the following: -Sets up a project to be your assistant -3 Daily scheduled briefings (morning, midday, evening) that keep you updated on email, calendar, and project status -Handoff process, in which each brief preps the next one. -HTML Dashboard to visualize Status and To-Dos I'm looking to create some content around this but need a few people to test it out for me, please reply or shoot me a DM for the prompt.

I'll build on this, I mentioned leveraging scale in my last post. One of the biggest unlocks with Claude Code and similar agents is the ability to work with data at scale. In real estate, what kinds of data sources and streams can you access to harvest for insights and intelligence? Public databases, industry news, trade publications? Start with some research to see what you can discover. An open ended starting quest like this will your brain going with ideas.

Ahh much better context. Primarily, I use these agents for marketing. They help do most of my social media and SEO. They analyze and optimize my ad campaigns. They also write email promotions for me. As a real estate guy, you might consider using AI tools to become a website flipper. It’s like digital real estate. You can get AI tools build and market the sites, then flip each for 4-6 figures (depending on niche and amount of marketing).

I wish it was actual magic, but the making money part is on you, and how you deploy these tools and workflows to reduce friction, increase productivity, and leverage scale. We'll open those doors, wide. I can't wait to roll up my sleeves for this. Don't get hypnotized by the hype merchants, no agent is magically making money while you sleep.

@IMJustinBrooke How do we make money with these agents? @TheViableEdge ?