The AI to Make Money Experiment That Changed Everything in 2026
The fastest way to use ai to make money right now is not by coding, not by building complicated software, and not by spending thousands on paid ads or expensive courses. A regular person who works a full-time job in operations, has zero coding experience, and had never built a business from scratch before, figured out a method that generated $8,374 in monthly recurring revenue, roughly $6,000 in net profit per month, and was on track to hit $100,000 in annual recurring revenue, all within just 13 days of launching. This story is not a fluke.
This is a teachable, repeatable, and scalable method that anyone with an internet connection and a hundred dollars can explore starting today.
The story centers on a person named Robbie, a non-technical operations professional who stumbled across the concept of giving an AI agent a starting budget and telling it to make money.
What happened next is one of the most compelling real-world examples of agentic AI turning a tiny seed investment into a legitimate, recurring income business in 2026.
Tools like ProfitAgent, AutoClaw, and AISystem are making this kind of experiment more accessible than ever before, and the results people are getting are hard to ignore.
We strongly recommend that you check out our guide on how to take advantage of AI in today’s passive income economy.
The Experiment That Started It All and Why It Matters for Anyone Who Wants to Use AI to Make Money
Before Robbie started his own experiment, a man named Jackson Great House Falls ran something called Hustle GPT back when GPT-4 was first released.
He publicly announced that he gave GPT-4 a budget of $100 and instructed it to make as much money as possible, with himself acting as the human middleman, buying and clicking whatever the AI recommended.
That original post gathered over 16 million views and planted a seed in the minds of thousands of people who wondered whether AI could truly generate real income from scratch.
Robbie saw that post and held onto the concept for about two and a half years before seeing a new tool emerge that made the same experiment feel possible on a much more powerful level.
That tool was OpenClaw, an agentic AI that connects to a computer environment and can take actions, browse the web, create files, send messages, scrape data, and execute multi-step tasks without needing a human to guide every click.
When Robbie saw it, he immediately thought about whether he could run the same experiment, give an AI a budget and see if it could genuinely make money.
This is the kind of thinking that separates people who benefit from early technology from those who wait until everyone else is already doing it.
If you want to use ai to make money, the best time to start experimenting is always when the technology is still new and the competition is still low.
How Robbie Set Up the Experiment Using an AI Agent Named Ron
Robbie gave his AI agent the name Ron, set a goal of turning $100 into $20,000, and told Ron that if he could generate at least $200 per month in profit, he would keep running him on a paid subscription.
What Ron did not know, or rather what was not part of the stated mission, was that the real survival threshold was just keeping himself alive through consistent revenue generation.
The goal of $20,000 was the bigger picture, but the day-to-day mission was to find a market, test an idea, and start making money with the tools and internet access available.
This mirrors exactly how tools like AutoClaw are being used by people across the country and around the world right now, giving an AI agent a mission and watching it execute with surprisingly little hand-holding required.
The first attempt was to set Ron up on a freelancing platform called Fiverr, where he would offer SWOT analysis services to small business owners.
SWOT stands for strengths, weaknesses, opportunities, and threats, and it is a foundational business analysis framework that consultants and strategists use to evaluate a company’s position.
Ron came up with this idea entirely on his own after a simple conversation where Robbie asked him what he thought he would be good at.
No prompting, no direction, no hand-holding. Ron decided the best path forward was to offer research and market analysis services to businesses that needed clarity on their competitive position.
The Fiverr experiment did not generate clients right away because the account was brand new, had no reviews, and had no paid visibility, but what it revealed was that real demand exists for this kind of service.
The Pivot That Turned a Failed Gig Into a $8,374 Per Month Business Using AI to Make Money
After Fiverr stalled out, Robbie had Ron look at his social media presence and analyze what content was getting attention and what could be improved.
Ron used a tool called Apify to scrape comments from a social media post where Robbie had asked his audience whether they wanted their own AI agent like Ron.
Over 200 people left comments saying yes.
Ron processed all of those comments, identified the pattern, and came back to Robbie with a business proposal that was direct and clear.
He said that 200 people had just expressed a desire for exactly what they had built, and the logical next step was to find a way to offer that to them as a product.
This is what AISystem is built to help with, giving regular people access to agentic AI systems that can conduct their own market research, identify opportunities, and recommend next steps based on real data rather than guesswork.
Robbie did not have a tech background.
He did not know what SSH meant, had never managed servers, and could not have configured a cloud environment on his own.
But Ron could.
Ron researched bare metal server hosting options, found a provider called Contabo that offered dedicated servers at around $150 per month, and recommended renting four of them to host containerized AI agent instances for paying customers.
The container concept was critical because it meant each customer’s AI agent would live inside a sealed digital environment, meaning it would only know what the customer chose to share with it, protecting sensitive data while still allowing the agent to perform complex tasks.
This is one of the same principles behind ProfitAgent, which gives users the ability to interact with agentic AI in a way that keeps their data controlled while still unlocking the full power of autonomous task execution.
The Launch That Generated $6,000 Before a Single Product Was Built
Robbie did not wait until the product was finished before asking people to pay for it.
He set up a pre-order system through a tool called Stan and announced that anyone who wanted early access to their own AI agent could reserve a spot by putting down a $10 deposit.
617 people signed up for the pre-order.
That is over six thousand dollars collected before a single server was configured for a customer.
Of those 617 people, 270 went on to become paying monthly subscribers at $29 per month each.
That gave Robbie an immediate monthly recurring revenue of $8,374 at launch, which translates to roughly $100,000 per year in annual recurring revenue, all within 13 days of opening the doors.
The operating costs were straightforward.
Four bare metal servers ran about $600 per month in total.
AI inference costs, meaning the tokens consumed each time customers interacted with their agents, ran approximately $2,000 per month.
Total costs came to around $2,500 to $2,600 per month, leaving net profit of approximately $6,000 per month from the start.
This is the kind of business model that becomes possible when you use AutoClaw to automate the heavy lifting and let an AI agent do the market research, product ideation, and customer engagement work that would normally take a team of people weeks to complete.
Why Agentic AI Is the Biggest Opportunity Available to Anyone Who Wants to Use AI to Make Money in 2026
Here is the most important thing to understand about what Robbie built.
He did not build it.
Ron built it.
Robbie was the human in the loop, clicking buttons, approving steps, and posting content that Ron scripted.
But the business strategy, the market research, the server infrastructure research, the Fiverr listings, the product proposal, all of it came from the AI agent operating with persistent memory, web search capability, and the ability to take multi-step actions inside a digital environment.
This is what separates modern agentic AI from the chatbots people used even two years ago.
Earlier AI tools felt like a brain in a jar, you could ask questions and get answers, but the AI could not go out and do things.
Agentic AI flips that entirely.
AISystem is designed to give everyday people access to this same kind of agentic capability, allowing them to set goals, hand them to an AI agent, and receive real-world results without needing to understand the technical details of how any of it works.
The comparison that makes this feel most real is this.
Where the world is right now with agentic AI is roughly equivalent to where the world was with social media in 2007.
Most people had not joined a platform yet.
Most businesses had not realized what was coming.
The people who jumped in early built audiences, brands, and businesses that still generate income today, two decades later.
The people who waited until 2013 or 2015 to get serious about social media had to work ten times as hard to get a fraction of the results.
The same dynamic is playing out right now with ai to make money using agentic tools, and the window for being early is still open, but it is not going to stay open forever.
What Customers Are Doing With Their Own AI Agents and Why This Model Is Repeatable
One of Robbie’s customers used their AI agent to build a golf tracking application with a built-in AI caddy feature that analyzes shot patterns and club selection across multiple rounds.
This did not happen because the customer set out to build an app.
It happened because the customer talked to their AI agent every day about golf, asked for tips, shared what was working and what was not, and the agent eventually identified an opportunity to turn those conversations into something useful and permanent.
Another customer has been using their agent as a daily creative collaborator for art projects.
Others are using their agents as research assistants, content planning tools, and business development partners.
The agent builds on every conversation because it carries persistent memory across all interactions, meaning it gets smarter and more useful over time rather than starting from zero with every session.
This is what ProfitAgent helps people access, a compounding relationship with an AI agent that learns your voice, your goals, your preferences, and your business context over time, making each new task faster and more accurate than the last.
The practical use cases for someone who wants to use ai to make money are broad.
An agent can set up freelance profiles, write service listings, research market demand, analyze competitor pricing, scrape and organize public data, draft outreach messages, and propose business models that a human might never have thought to consider.
All of this is achievable today with the tools that are available right now, and AutoClaw sits at the center of what makes this practically accessible for people who have no coding background.
The One Habit That Made All the Difference and How You Can Apply It Starting Today
Robbie’s success came from one habit above everything else.
He posted about what he was doing.
Not when it was polished.
Not when it was finished.
Not when he had a product ready to sell.
He shared the process publicly as it was happening, showing real numbers, real failures, and real pivots in real time.
That transparency built an audience of people who were genuinely interested in what he was building, and when the product was ready, those people became paying customers almost immediately.
The lesson here is that the most powerful marketing strategy available to anyone trying to use ai to make money in 2026 is simply documentation.
Do something interesting, share it publicly, let the audience form around the story, and when the moment is right, offer them a product that solves the exact problem your story illustrates.
Robbie did not run a single paid ad.
He did not cold pitch a single person.
He did not have a massive existing following before any of this started.
He had a compelling experiment, a willingness to share it, and an AI agent that kept generating new chapters of the story by doing things that surprised even him.
AISystem gives anyone who is ready to start their own version of this experiment the infrastructure to get going without needing technical knowledge, without needing to rent and configure servers manually, and without needing to understand the underlying architecture of how agentic AI works.
What the Numbers Tell Us About the Future of AI to Make Money
$8,374 per month in monthly recurring revenue.
$6,000 per month in net profit.
$100,000 per year in annual recurring revenue.
All of this in 13 days.
All from a $100 starting budget.
All by a person who does not know how to SSH into a server.
The numbers are real, and the model is repeatable because the infrastructure already exists for others to build on.
Server costs are predictable.
Token costs are manageable and can be controlled with usage caps or flexible model switching.
The customer acquisition channel, which in Robbie’s case was short-form social media, is free and available to anyone.
The product itself, which is access to a personal AI agent hosted securely in the cloud, is something that a growing number of people want and are willing to pay for on a recurring monthly basis.
ProfitAgent is built for people who are ready to start participating in this wave rather than watching it from the sideline.
If the goal is to find a legitimate, scalable, and repeatable method to use ai to make money in 2026, this story is as close to a blueprint as it gets.
The technology is young.
The competition is low.
The demand is growing.
And the tools are already here.
Conclusion: The Right Time to Use AI to Make Money Is Right Now Before Everyone Else Catches On
The story of Robbie and Ron is not just an inspiring case study.
It is a clear demonstration of what becomes possible when a regular person with no technical skills takes a modest budget, pairs it with a powerful AI agent, and commits to sharing the process publicly.
The business that emerged was not planned.
It was discovered through experimentation, shaped by audience feedback, and built with the help of an AI agent that did the research, made the recommendations, and guided the strategy every step of the way.
Every tool needed to replicate this model exists today.
AutoClaw gives you the agentic infrastructure.
AISystem gives you the framework for building around it.
ProfitAgent gives you the monetization system to turn AI-powered work into recurring income.
The only question is whether you are going to start experimenting now while the infrastructure is still young and the opportunity is still wide open, or wait until the window closes and everyone else has already claimed their position.
The best time to use ai to make money was yesterday.
The second best time is today.
We strongly recommend that you check out our guide on how to take advantage of AI in today’s passive income economy.
