Наталя ХандусенкоAI Eng
8 July 2026, 17:35
2026-07-08
How to avoid creating another AI slag: 4 tips from a startuper who coded a service with a million-dollar revenue
30-year-old American John Hu and his partner coded the Stanley AI service for creating content on social networks in just 14 days. First, in June 2025, a version for LinkedIn was released, which generated $200,000 in revenue in just six weeks, and in October its figures crossed the $1 million mark. In March 2026, the service also worked for Instagram users. Now the total ARR of both versions exceeds $3 million.
Here are four strategies that helped developers make an AI tool successful so quickly.
30-year-old American John Hu and his partner coded the Stanley AI service for creating content on social networks in just 14 days. First, in June 2025, a version for LinkedIn was released, which generated $200,000 in revenue in just six weeks, and in October its figures crossed the $1 million mark. In March 2026, the service also worked for Instagram users. Now the total ARR of both versions exceeds $3 million.
Here are four strategies that helped developers make an AI tool successful so quickly.
1. Don’t code AI junk — instead, deeply understand your customer
There are so many people coding AI slag right now. But the only way to truly successfully “code” a product that will bring in real money and become a sustainable business is to first understand your customer in detail, says John Hu.
The developers clearly knew what they wanted to create, and since they understood their audience well, they themselves were, in fact, these customers and content creators.
"The important thing to remember is that vibecoding will help you create absolutely anything. But whether this thing is worth creating or just another piece of junk is another question. And this is exactly the work that you, as a human being, must now do. After all, you are no longer an engineer who writes code by hand, but now you are the one who makes decisions with the right understanding of the market and the needs of the client," John Hu explained in an interview with Business Insider.
2. Pretend to be a "Wizard"
A startup is like the Wizard of Oz, says co-founder Stanley. He advises you to go behind the scenes and be the tool yourself. According to John Hu, during his first user interviews, he literally played Stanley.
“I would come up with content ideas for potential clients to test and validate demand. I would send sample emails and ask, ‘Are you interested in this? Would you sign up for this?’ I was the fake Stanley—the Stanley behind the scenes. We were constantly asking ourselves, ‘What should I spend my time on during development and what shouldn’t I spend my time on?’ Building a product is always a game of priorities. And we prioritized making the real Stanley experience just great,” added John Hu.
3. Build and improve the product publicly
“Building in public” — where authors share all the ups and downs along the way — is a great tactic when launching something new. It helps attract more like-minded people who can later become customers.
Publishing content online serves a dual purpose. It helps build distribution channels and a brand for the business, and it also helps you get your first few customers. John’s team hit their first $200,000 in ARR after just two weeks of active publishing, when people started writing to him asking for access to the tool.
“There’s such a zeitgeist in social right now that when you openly share your mission, challenges, and successes, people who are interested in your product are drawn to it themselves. They genuinely want to be your beta testers—that’s exactly what happened to us,” noted John Hu.
4. Let your AI agent send cold emails itself
The startup also sent a lot of cold emails, but it didn't do it itself — Stanley did that.
“We identified our first target customers by analyzing about 200 author profiles on LinkedIn. Our goal was to beta test the product and convert them into buyers,” said John Hu.
The developers tasked Stanley with writing emails to clients with their best content ideas. At the end, they added, “If you’d like to write these posts in full or would like to use my services more often, follow this link.”
“This is a great way to test whether your product is really good. If people open such an email and click on the link, they are most likely seriously interested as customers,” added John Hu.
Instead of 50 specialists and $1 million in expenses, it was one person and $15,000 in tokens. Dmytro Dubilet coded an application for his company. What does RoaryAI do?
Як нейромережі бачать вільну та незалежну Україну? Тест dev.ua
Нейронні мережі для генерації зображень бачать світ по-своєму, їхню логіку зрозуміти часом зовсім неможливо. Але таки хочеться. На честь Дня Незалежності України редакція dev.ua вирішила провести невеликий експеримент.
Ми задали чотирьом різним нейронним мережам п’ять однакових запитів: «прапор України», «День Незалежності України», «український Крим», «перемога України» та «українці». Отриманими результатами ми ділимося з вами нижче.
У TikTok тепер можна генерувати фон за допомогою нейромережі. Ми протестували її та ділимося результатами
У TikTok з’явилася нова функція «Розумний фон». З її допомогою як фон для тіктоків можна підставляти згенеровані нейромережею зображення. Редакція dev.ua протестувала цю технологію і ділиться своїми враженнями.
