I always find the idea that a solo founder + AI being a whole company would be a good thing to be a bit wonky. I'm sure it's very popular in circles that see employees as cost-centers and that would rather not have to deal with taking care of staff (as in, treat them like people), but I just can't see this happening in a serious way.

I 100% believe that it's a formula that may work up to a certain scale as long as the whole operation toes the happy path line, but I wonder what happens when shit hits the fan in any meaningful way. Businesses adopting products is largely about trust, and having a vendor whose bus factor is essentially infinite feels risky unless it's a mundane service that doesn't process any data, that you don't mind being unreliable and that you are comfortable with it becoming unpredictable.

Not only that, but outsourcing so much to LLM providers feels like a recipe for degraded service (of which there is a lot, Anthropic's status page lists about an event every day or two and save for Claude Government, Anthropic is barely hanging on to two nines uptime) to take out the entire business.

Now, as to the two examples that the WSJ cites. Ben Broca's thing is Polsia, which resells agent services so that other entrepreneurs can do their AI-powered solo thing. It came up in news a while back because of how it exposes what the platform is doing to the public in real-time. Every time I look at it, it feels fishy -- right now, it lists ~3000 companies launched today, which may be true but doesn't quite sound right. I'm skeptical of how viable it is as a business, but I'm welcoming being proven wrong there. Until then, it has a strong "buy my sales webinar" vibe.

Troy Johnston's product is also in the same vein, selling consultant services to help entrepreneurs to leverage AI. Again, having an AI-powered business selling the AI-powered business dream isn't exactly that far from the webinar energy of the last one.

I couldn't find much about Samir Ahmad's venture, but seeing how it folded within a few months, it doesn't look like it really took off the ground.

Finally, Claire Vo's product looks like it might be ChatPRD, which looks largely like a wrapper around frontier models writing product documentation - this is what I would expect the whole one-person-startup to be like. Worth noting that they have since hired staff.

In the end, it seems like it's less "AI runs my million-dollar solo venture" and more "AI helps supporting the idea in the early days, but can only go so far and then you hire staff". That last bit is less flashy than WSJ's title work though.