The Startup Slop Problem
Infinite Instant Businesses
A classic scam
Imagine you get an email telling you Nvidia is going to skyrocket over the next week. Lo and behold, it does. A week later, the same account sends you another email saying Google’s gonna rocket. It does! You then get a third letter saying, “I was right with the first 2, but I have a third stock that is really going to hit… gimme $1000, and I’ll tell you what it is.” Well, if they nailed the first two, you’d be a fool not to fork over the money for this hot tip. Right?
Of course, this is all a ruse, known as the Baltimore Stockbroker Scam. The trick you aren’t aware of is that you weren’t the only person to receive these emails. Thousands of people got it, with variations for many companies going up or down each week. Most of them were wrong and were ignored by the recipients. But just like blindfolded monkeys throwing darts at stock tickers, the law of large numbers says sometimes you’ll pick the right one and “beat the pros.”
You didn’t know about all those failures. You only knew about the one success that you received. And in the world of scams, the one that gets you is the only one that matters.
Quantity brings quality
If you throw enough independent solutions at a problem, you’ll likely find something that fixes it. This is a closely related cousin of the concept that the best ideas come from LOTS of ideas. Famously, Jerry Uelsmann conducted an experiment at the University of Florida in which one group of students was asked to be selective and take the best-quality photos. Another group was told to take as many photos as possible. In the end, it was the “quantity” group that also produced the best quality. You take more shots on goal, and eventually you’ll score more goals. We put constraints on being right because, in life, we have constraints. We only have so much time, bandwidth, and patience to run and evaluate these scenarios. But in a world where AI is cheap and ubiquitous, the old math isn’t applicable.
Instant business
This takes me to a new AI agent tool I signed up for last week. The name of the company is Polsia (which slyly is an anadrome for “AI Slop”). Polsia promises to be the “AI That Runs Your Company While You Sleep.” I logged in for a free three-day trial ($50/month thereafter), and it scoured the internet to learn everything about me. Then it spun up a business that it thought I’d like to own. I want to emphasize that I did nothing other than click “OK” a few times; the AI did all the work. In minutes, it created a company called Moneta with a nice-looking landing page and a decent-sounding objective:
Moneta is the AI agent that monitors every transaction, catches fraud in real-time, fights chargebacks autonomously, and routes payments through the cheapest path. Your entire payment ops team — replaced by one agent.
Behind the scenes, it did real work too. It built a pretty, but rudimentary, monitoring dashboard, so there was some semblance of a product working behind the scenes, and it also started doing social media and cold outreach. Again, I was doing nothing other than watching daily status emails come through.
Now, to be clear, it didn’t make any sales in the 72 hours since launching, but it was able to build a website, a back end, and integrate Stripe. (I’m glad I didn’t make sales. Honestly, I have no idea if it could have fulfilled the promise, and I didn’t want to find out at a customer’s expense.)
Doing this one time is a party trick, but doing it 10, 100, or 1 million times is when I start to get intrigued. I don’t know the economics of Polsia (they do claim $1.5M ARR already), but I do know the clear trend for using AI services. Token prices (i.e., the cost of running these models) have plunged logarithmically and show no signs of stopping. When the economics of something tell you it will be 99% cheaper to do the same thing in a couple of years, it opens up many possibilities. Soon, we will be able to spin out millions of these companies without cost or effort. And that will be a problem.
Death of solopreneurship?
I have been preaching for years that the job of the future is a solopreneur. Solopreneurship was my vision of the democratized future of work. You have one person, hyper-focused on solving a problem, augmented by AI, building real things that customers really need. But Polsia reveals something troubling: you don’t even need a human for a business. The “founder” is nothing more than a needed signature for filing a Delaware C Corp.
If there’s no human intention behind these businesses, what are they, then, other than noise crowding out the legitimate founders with a vision? I previously wrote about how we’ve been tackling the problem of mass-produced media for over 500 years. No one today complains about the technology of the printing press. That enabled more voices to be heard, and this will enable more experiments, which is a good thing. But will it overwhelm? How much will be too much?
This brings us back to the scam that began the piece. The Baltimore Stockbroker didn’t need to be right every time; he just needed to flood the zone and wait. Polsia’s zombie companies work the same way. Most of them, including Moneta, will die quietly, unloved and un-Googleable. But one will land a customer, get a positive review, and probably raise a $20 million seed round. And the meatpuppet who signed the articles of incorporation will have been asleep the whole time.
We’ve always had startup graveyards. We’ve never had startup factories that run while you dream.
Tremendous experiment, Mike! As always.