How Did a Company 'Producing AI Garbage' Raise $30 Million?
Deep thoughts on AI and aspirations —— ByteThink Circle
A company founded just six months ago is being called out for “producing AI garbage,” yet has just raised $30 million. The critical reports list a litany of damning data: terrible retention rates, the vast majority of products abandoned after launch, and so-called revenue that’s half advertising.
By conventional logic, this company shouldn’t survive. But capital is rushing in. This tells us one thing: the controversy itself may be closer to the truth than those pretty numbers.
The Real Meaning of the Controversy: Can AI Actually Operate a Company Independently?
What this company does is extreme—you give it a startup idea, and it develops the product, runs operations, buys ads, finds customers, and makes money. Humans step aside while AI takes the wheel.
This hits on a real question the entire industry is avoiding: Can AI truly run a company independently from start to finish? Not just improving efficiency at certain stages, but orchestrating everything end-to-end. This company is answering that question in the ugliest way possible (massive failed attempts, obvious garbage output).
So capital is willing to invest—not because of current earnings, but because of the answer to the question it’s pursuing.
An Undervalued Signal: Real Data
This company is actually more honest than most AI companies. It doesn’t hide the garbage or embellish retention. It openly admits most products don’t work out, and offers a crucial statement: “The quality of the garbage improves every week.”
This line deserves closer thought. A large volume of cheap failed attempts is itself a form of exploration—using extremely low costs to filter out the few viable paths. This is completely different from companies betting big on one direction. The former is systematically finding answers; the latter is gambling.
And it has a very tangible revenue metric: users are willing to invest in advertising on its platform. Users only keep spending money if they believe this product can make them money. This signal is more real than any GMV—because it reflects users voting with real money.
Bubble and Real Value Must Be Viewed Separately
Of course, there’s definitely a bubble component here. How much of the $30 million valuation is real business versus story and hype is unclear. The vision of AI independently operating companies is still far from mature, and many will die along the way.
But bubbles and value often coexist. The key is whether there’s a real signal underneath to support it—and for this company, that support is the signal that “users are willing to invest in ads.” As long as this signal holds, the story has foundation; when this signal drops, that’s when it truly collapses.
Judgment for Entrepreneurs
My assessment: This company’s greatest value isn’t itself, but the question it raises.
If you’re building AI applications, two questions are worth considering: First, does my product have a signal where users vote with real money? Second, am I systematically exploring directions in a cheap way, rather than betting on one direction?
These two questions are far more useful than fixating on this company’s fundraising numbers.
Key points: The controversy points to a real question (Can AI independently operate a company?); “Garbage quality improving weekly” signals systematic exploration; users investing in ads is a genuine metric; bubble and value coexist—look for real signals underneath.