Lately I've been thinking about AI and the rise of one-person companies, especially in China.
For the first time, it feels plausible that having an idea might be enough.
Design, development, testing, deployment, documentation, and even marketing are increasingly tasks that AI can assist with. What used to require a team may soon be achievable by a single person.
Historically, turning an idea into a product involved a lot of friction. Ideas were debated, challenged, refined, and often abandoned. Building software required time, money, and coordination. Those constraints acted as a filter.
Many ideas never made it into production. Often for good reason.
Today that filter is disappearing.
A solo developer can build and launch products that would have required an entire company just a few years ago. That creates enormous opportunities. It also raises an interesting question:
What happens when the cost of creating software approaches zero?
My concern is not whether we will be able to build more software. We clearly will.
The question is whether we will be able to find anything worth using.
We have already seen something similar happen on the web. The early vision of the internet was a world of knowledge, discovery, and meaningful digital experiences. Instead, we ended up with an overwhelming volume of content competing for attention.
You can see the same dynamic in media consumption today. Many people spend most of their online time on TikTok, YouTube, or Roblox. Long-form content, books, and deeper forms of learning often struggle to compete with highly optimized streams of entertainment.
Software may follow a similar path.
If creating an app becomes nearly effortless, the challenge will no longer be building it. The challenge will be standing out from millions of similar products, all generated faster than users can evaluate them.
Perhaps the scarce resource of the future will not be engineering talent, but attention and trust.
I'm curious what others think.
Will AI lead to an explosion of innovation, or are we heading toward a world flooded with software that nobody really needs?
Top comments (1)
I think software creation is becoming cheaper, but distribution and trust are becoming more expensive.
A few years ago, the bottleneck was building the product. Today, many people can build something in a weekend with AI assistance.
The harder question is: why should anyone use it?
When creation costs approach zero, quality, reputation, community, and distribution become the real differentiators. We may end up with millions of apps, but only a small percentage will earn enough trust to get consistent users.
In that sense, I don't think engineering disappears. The scarce resource just shifts from code generation to problem selection and trust building.