This story is part of CNBC does this The Moment series, where successful people reveal the critical moment that changed the trajectory of their lives and careers, discussing what inspired them to take the leap into the unknown.
Dhruv Amin and Marcus Lowe left Google and created a startup that was “profitable from day one,” operating at a rate of $2.2 million by September 2023, Amin says. They predict continued success, forecasting that monthly revenues will increase at least fivefold by the end of 2024.
Then they shut everything down to start from scratch – all because of ChatGPT, says Amin.
Today, Amin and Lowe, both 33, are the co-founders and co-CEOs of Anything, a ambiance coding startup valued at $100 million following a $11 million funding round in September, according to the company. The company was originally known as Create, a marketplace connecting startups with independent engineers and artificial intelligence-powered coding tools to build websites and apps.
Why blow it up? Amin cites “the ChatGPT moment” in November 2022, when OpenAI launched the generative AI chatbot. Its leap in capabilities over previous AI models “was a surprise to everyone,” Amin says.
DON’T MISS: The Ultimate Guide to Starting a Business: Everything You Need to Know to Be Your Own Boss
By spring 2023, Amin and Lowe predicted that AI could eventually approach the ability of human developers to write sophisticated code. If this happens, human engineers may no longer be needed, making their business obsolete.
But if they rebuilt their business around AI and the technology didn’t advance fast enough, their new model would be a failure. Customers would flee and the business would run out of money.
After months of anxious deliberations, in October 2023, Amin and Lowe began the difficult process of laying off half of their seven-person team and cutting ties with independent developers, Amin explains: “Within two weeks, we were back in an empty office.”
They first launched an AI-powered tool to create application components such as login forms and calendar tools. In April 2025, they launched a new product for building entire online businesses, including back-end authentication and payment systems, no coding experience needed – and renamed the company Anything.
“That’s actually when I felt like it really took off,” Amin says. Within two weeks of company arrival relaunch announcementIt all grew to an annualized revenue run rate of $2 million, he says.
The AI coding industry is “extremely early” in its development, says Amin, as evidenced by overall mixed reviews for ambiance coding services. Yet non-technical users have created working business applications with Anything, including a hair salon owner’s AI stylist mobile app and a dental hygienist’s app to track gum health, Amin wrote in a blog post on September 29.
Here, Amin talks about realizing that AI could upend their entire business model, the anxiety of whether to start again, and good timing when launching a startup in an emerging industry.
CNBC Make It: You decided to close your doors and rebuild in October 2023. How painful was that decision and what was the reaction from your company’s investors?
Amin: We came back to zero, so it was hard. We had to let go of half the team. We had to close the developer marketplace and tell our customers that we weren’t going to do this with developers anymore.
Back then, people were still laughing, “Is this AI thing going to be real?” Are you really going to be able to create real applications (with only AI)?
We had raised ($3 million) in venture capital. We promised (investors) that we would meet (specific) deadlines. Then we went back to everyone and said, “Actually, no, we’re going to go back to zero.” Everyone was like, “Are you sure?” »
As you and your co-founder followed the advancements in AI, what were your conversations like?
We talked all the time. We reviewed the position about once a month. It took Marcus and I a very, very long time to come to a conclusion (to restart the company), and it felt like one of those decisions that we as co-founders both had to agree with.
There were times when I was more optimistic about the prospects of the (original) market and what we could do with it. There were times when Marcus was more optimistic. And there were times when we were both like, “This technology is advancing so quickly and we’re using it in more and more places, does it make sense to create a new product?” It was hard.
I think what ended up breaking that logjam for us was: We launched a few prototype builders, just to test and see if there was a demand for (code generation products) where you don’t have to talk to any of us, you can just use it. This gave us a first signal. When we talked with our customers, we felt like the code generation system was right for the product and the market.
How confident were you that the underlying generative AI technology would continue to advance exponentially?
Confident, but not confident about deadlines. That was the scariest part. Until 2024, the major language models have sort of reached a plateau, to some extent, in terms of capacity. At that time, we didn’t really know how good (this technology) would eventually become. There were definitely (times) where we were like, “Oh, boy. Did we make a mistake trying to pivot?
Can we even build something valuable and useful enough that people will want it and pay us for it? We had a lot of opportunities to pivot to a quicker win, which would be, “Let’s create a prototyping tool. Let’s create a design tool, because all (LLMs) can do is components. We’ll sell it to agencies.”
What stopped you?
Marcus and I always had this (mindset) that we had already gone through a really big turning point in trying to align more closely with the company’s mission. This pivot would have been for nothing if we hadn’t actually accomplished the company’s mission, which is for people to be able to create real production products, even if they don’t know how to code.
Now that’s a useful founding story: this company will get through this, even if the technology changes, even if the business model changes. At the time, it was hard. Now I can say with some pride: you don’t have to take my word for it that we are serious about what we are trying to do here. We painted it with blood, sweat and tears.
I thought it would take about a year for this technology to be good enough to build real production systems. It actually took about two and I still think we’re just at the beginning. The sophistication of the apps you can create with Anything – what they look like a year from now, or even two years from now, will be dramatically different.
Hopefully our hardcore kingpins are behind us at this point. Now I think we’re just more in the (mindset): let’s execute, take advantage of the opportunities and really develop them.
Where would your business be today if you hadn’t closed the market to focus on AI?
It’s so hard to say. I wonder if the company would still exist? Would the company we were building then still be attractive to (investors)? I think market models like this, while popular in 2021, have really gone out of fashion unless you are more focused on pure AI.
This conversation has been condensed and edited for clarity.
Do you want to be your own boss? Sign up for CNBC’s new online course, How to Start a Business: For New Founders. Discover step-by-step tips for launching your first business, from testing your idea to growing your income.