He Turned $400 Into $2.5M Using AI

[00:00]

12 bucks an hour and you can afford it. It opens it up to anyone. This make an app in like 3, 4, 5 weeks without the big investment. I wonder if I could vibe code a piece of software and have my first customer by the next day. And I ended up getting my first customer by the following Tuesday, a $15,000 customer. And then from there it blew up. Not at all. Stop assuming that other people know what you know. They don't. If you're listening to this right now, you're in like the 5%. How did someone get their first customer? Whether it's pretty simple. I just said I'm going to start calling businesses. And over the next six weeks or so, I ended up with about $180,000. We're looking at about $8 million in recurring revenue by the end of this year. After one year, and you started this with about $400. The big thing here is that I don't have anybody to answer to. I am literally getting on a plane in five days. And it's okay because I'm the boss.

[00:00 Continued]

Business that helped other business owners save basically all of us know how to do. And within and over a million in net profit. And he didn't code. This year, 2026, he's on pace to do over $7 million with three or four employees. This is an incredible story. John, how did you do it? How can we copy you? John, why don't you tell us who you are and what you do? The founder of a company called the General AI Institute. A little bit easier to say, but that's really what I'm doing: helping people use AI, learn how to implement it, and really try to transform their businesses.

[00:00 Continued]

When did you start this? A little over a year ago. February 27th last year. I'd been vibe coding a bunch and I just woke up and said, "I wonder if I could vibe code a piece of software and have my first customer by the next day." It ended up taking me instead of two days, I ended up getting my first customer by the following Tuesday. It wasn't just like a little tiny one. I was a $15,000 customer. And then from there it blew up. And you're not a coder. I couldn't write a line. If you said, "John, can you write me one line of code?" I would be like, "No."

[00:00 Continued]

When I first sold my company, I dove in. I had this idea in my mind. I wanted to start a travel social network, like a travel log. I have on my wall a pin board of every place I've visited in the world. I wanted to create a digital version of that so people could see where someone had been. I contacted developers from my previous company and asked for a quote. They said it was going to be $105,000. That was going to give me iOS, Android, and web.

[00:00 Continued]

I hung up with those guys, jumped on LinkedIn, and saw somebody saying they built software in like 10 minutes in front of me. I said, "What's a Replit?" I had no idea what that was. It said I just had to describe what I wanted. I took this $105,000 plan, put it in there, and said, "Build this." I dragged it in. Twenty minutes later, it was built. And I was like, "Whoa." I started working on that and got addicted to it.

[05:01]

Being out here, I think there are less techy people out here. They're like construction companies, blue-collar businesses. Whether I was at church or out to dinner, they just knew nothing. I was like, man, I should do something about it. That's really what I wanted to do: help people see this amazing technology and help them improve their own business.

[05:01 Continued]

You get a $105,000 quote, which was probably the cheaper quote. You take the proposal and learn about the magic of vibe coding. The idea was vibe coding with AI for business owners. When I discovered that to be able to go and pay this dev shop $105,000, I instead needed like $30 to try Replit. That change was so different. $100,000 to $30. For the current business, it took me about three days. It was an intense three days, 10 to 12 hours a day: buying the domain, setting up my email, setting up different things. All in, about $400.

[05:01 Continued]

I knew the point it took me three days to get to would have required developers for a long time. I did the calculations. The salaries and everything would have cost me about $3.2 million. It cost me $400. That lit me up. It was a frame-breaking moment.

[05:01 Continued]

Chris reads a snippet from an email he sent Jon. He argued that Jon had something with a high chance of making millions in the foreseeable future. He contrasted that with the much lower probability of a new idea making life-changing money, especially if it required raising capital and dealing with liquidity preferences. He advised staying focused on the current opportunity.

[10:03]

Jon agreed with the email. Most people don't get liquidity in venture-backed businesses. In his current business, he has nobody to answer to. He can get on a plane to Paris with his wife because he is the boss. He is self-funded. He contrasts that with venture funding, where investors may push founders to continue until they get their money back. In the new world of AI, he argues, you do not have to do that. Someone making $12 an hour can use these tools and build something. It opens it up to anyone.

[10:03 Continued]

After one year in business, the goal for the year is close to $8 million in recurring revenue. At the time of the interview, they are doing about $400,000 per month in recurring revenue. The business was started with roughly $400 and became cash-flow positive quickly. Jon says he made as much money in this business in one year as he did after selling his prior business. The current business may be worth a six or seven times multiple because it has SaaS elements.

[10:03 Continued]

The first version of the business was an AI IQ test. The idea was to measure how AI-ready someone was, then sell courses and certifications. People could put the certification on LinkedIn or a resume. The first weekend, that was the whole idea: build an AI IQ test and sell the certification. Later, the business moved toward selling directly to businesses instead of individuals. Jon reached his first million around six months in, largely solo with AI tools and a few contractors.

[15:01]

Jon notes that Replit gave him credits after the CEO mentioned him publicly. He now has five full-time employees in addition to himself. They have real salaries and are delivering for clients. He expects around $3.5 million in profit by the end of the year.

[15:01 Continued]

In the beginning, he ran ads for the AI IQ test. Around 150 people clicked, but nobody purchased. He later discovered he had messed something up with Stripe. It took five minutes to fix. He had been posting publicly about the journey: "I started this business. I don't have any customers yet. This is what I'm doing." Then he began reaching out to people he knew and saying he had a new thing measuring how good companies were at AI.

[15:01 Continued]

After about six messages, one person said they had just been talking about AI as a team. They got on a call. Jon took him through the test and started helping him. When asked what it cost, Jon said, "15 grand." The client said, "Done." From there, Jon realized he should start calling businesses. Over six weeks, he reached about $180,000 in sales.

[20:00]

The original offer was measurement and training. Jon would assess the organization, get people together, and train them on AI: how to use different models, how to double-check work, how to get good answers. Many trainings were delivered over Zoom with 50 or 60 people. Clients loved it. At first, the model was one-time, but a client asked if there was a model where Jon would stick around. That sounded like recurring revenue.

[20:00 Continued]

Jon invented a Chief AI Officer service. He said, "Pay me 15 grand a month and I'll come in and do the training, but then we'll follow up and make sure this is actually impacting your business." He framed it around the possibility of increasing revenue by 10%. For the right client, $180,000 per year was small relative to the upside. He called a few other customers over the next couple weeks and closed half of them. Some signed up for $25,000 packages because they were larger companies.

[20:00 Continued]

Pricing started as a package of hours: $15,000 for 15 hours per month, $25,000 for 25 hours per month. Over time, the model became more system-driven: installing systems and tools that would stay inside the company. The goal is to be embedded in the company's ongoing AI infrastructure, not simply easy-to-cut hourly help.

[20:00 Continued]

Early clients were businesses without CTOs: law firms, construction companies, masonry companies, and similar organizations. They usually had good business processes but lacked technical leadership. AI allowed Jon's team to build processes and let AI take care of tasks that previously required people.

[25:03]

The company typically does not work with businesses below about $8 million in revenue. The average client revenue is around $30 million. The best fit is a company big enough to have complexity and budget, but not so large that it already has a sophisticated internal technology function.

[25:03 Continued]

Common use cases include CRM discipline, sales and marketing workflows, finance reporting, executive dashboards, and AI-enabled internal assistants. Jon describes using Slack channels for clients, AI systems that watch for tasks, CRM updates, report generation, and dashboards that give CEOs better visibility.

[25:03 Continued]

The sales playbook begins with a simple question: "What are you doing with AI right now?" Most businesses say they use ChatGPT to help write emails or summarize things. Jon reassures them that they are normal and still ahead of many companies just by thinking about it. Then he explains that his company helps in three ways: strategy, transformation, and education.

[25:03 Continued]

Strategy means helping the CEO understand how AI will affect the business and competition. Transformation means building actual processes and workflows. Education means training the team to think differently with AI. Jon explains that software can now be built for one task and thrown away, which requires people to think differently about what is possible.

[30:01]

Jon warns people not to become a dev shop, especially if they are not developers. If you promise to build every process for the client, you or someone else must maintain those processes long term. The best customers are people on the client's team who want to figure things out themselves, such as directors of operations or finance leaders.

[30:01 Continued]

Jon compares the model to coaching piano or gymnastics. If the coach does everything for the student, the student does not learn. But if the student learns a new skill, they feel excited and associate that progress with the coach. In the same way, when internal team members learn AI skills, they like the consultant more and want to keep them around.

[30:01 Continued]

The goal is to sit with a client team member for an hour if needed, help them build something, let them solve their own problem, then teach the next layer: Replit, an agentic framework, or another tool. Because AI advances quickly, the consultant can continue bringing new useful lessons.

[30:01 Continued]

When teaching beginners, Jon often asks who is a coder. Most people have never done it before. He opens ChatGPT or Grok and has them create something fun, like Tetris or a dinosaur game, just by talking to AI. This makes the learning process playful and tangible.

[35:00]

To get the first customer, Jon says there is no vibe coding shortcut. You have to reach out to people and talk to them. You can use AI tools to research businesses and find use cases, but then you must call them. He landed the first one, called more people, and kept doing it.

[35:00 Continued]

Jon emphasizes public posting. The best way to get no customers is to make sure nobody knows what you are doing. He argues that this is not necessarily about becoming an influencer. It is about telling people what you are building. Your Facebook friends, local network, and LinkedIn connections likely contain business owners.

[35:00 Continued]

He tells the story of a woman who took his "build a business in a weekend with AI" course. She told her son about it. The son gave himself a challenge to automate himself out of a job. Within days, he had automated enough of his work that he was afraid to tell his boss. His mother told him to tell the boss anyway. He did and immediately got a promotion to teach everyone else how to do it.

[35:00 Continued]

The lesson is to stop assuming other people know what you know. If you are in the AI sphere, you are likely in the top few percent. Many people do not know tools like Replit and only use ChatGPT casually. Assuming people know too much can prevent you from offering real value.

[40:03]

Jon discusses sales calls. Many deals are already warm by the time he gets on the phone. If someone has agreed to discuss AI automation, they are usually interested. The question becomes whether they hate the cost more than they hate their current problems.

[40:03 Continued]

The local aspect matters because trust matters. It does not have to be purely geographic, but shared context helps: same city, same school, same hobbies, same networks, or online connections. Posting about normal life can help people identify with you before they discover your AI work.

[40:03 Continued]

Jon expands the STE framework: Strategy, Transformation, and Education. Strategy helps the CEO stay ahead of competition. Transformation improves operations. Education trains the team. He explains to CEOs that AI can help them avoid unnecessary hires, build better systems, improve margins, and scale more effectively.

[45:01]

In the sales conversation, Jon asks what the prospect wants to automate or who they would like to hire. This gets the buyer to describe their own problem. Then Jon connects the service to that problem and frames the ROI. If AI can prevent a hire or increase productivity across existing employees, the monthly retainer becomes easier to justify.

[45:01 Continued]

Jon agrees with dropping small technical insights into the call. People buy from people they think are smart. He may mention APIs, systems, or integrations, not to overwhelm the prospect, but to show that he understands how the solution might work. Then he returns to the message: "This is why you hire us. We're the experts. We can do this."

[45:01 Continued]

Jon closes with encouragement. Compared to listeners, he says he has skill gaps too. AI fills in those gaps. You do not have to be afraid to ask AI. In earlier businesses, he learned enough from technical people to ask good questions. Now AI can act as a tutor or coach. In a couple days, someone can figure out enough to create useful value.

[50:03]

Jon says people can find him primarily on LinkedIn by looking up Jon Cheney. He also posts on TikTok, Instagram, and Facebook under Cheney Piano, where he shares piano, AI, and entrepreneurship content. He loves helping people and is happy to point them in the right direction. Chris thanks him and closes the episode.