Building in public is still a good strategy.
But in the AI era, it carries a very different kind of risk. In the past, sharing what you were making helped you attract fans, early users, and fellow creators. Now the situation is different. The problem framing you worked hard to refine, the copy, the notification timing, the UX flow, and even the story of why you built the product can all become training material for someone else.
I feel deflated when I see copies that mimic even the timing of alerts and the phrasing. What feels unsettling is not just that the features look similar. It is the feeling that my judgment and taste are being turned into a cheap reference.
Even so, I do not plan to give up on building in public entirely. I am simply moving toward hiding important work from the main blog feed, or showing it only to people who have the link. These days, public exposure is both marketing and free market validation, so some things still need to be visible.
The same kind of complaint keeps appearing on Hacker News, Reddit, and Indie Hackers. Vibe coding and AI agents have clearly improved productivity, but they have also increased the speed and shamelessness of copying other people's products.
The most striking example looks like this.
One developer open-sourced an enterprise RAG architecture they had refined for two years. Not long after, another founder reportedly showcased that repository as if it were a flagship output of their AI agent, explaining it to investors and customers as though their own system had produced it. What stood out was not just the anger, but the sense of emptiness. It becomes very easy to understand why someone would say they feel like the free R&D department for copycats.
Another developer shared a personal story on Reddit about why they built their app in the first place. A few days later, copycat apps appeared in the App Store with not only similar features, but also that personal story rewritten by AI and repackaged as if it were their own. Copying features already feels bad. Copying the origin story feels much worse.
That is what makes this different from before.
Old copycats mainly copied features. AI copycats now mimic features, copy, positioning, and founder narrative all at once. They do not just imitate the surface. They produce a shallow copy of the product's entire context.
The paradox of building in public
I do not think building in public is over.
It is still powerful in the early stage. Sharing the process brings feedback, attracts potential customers, and makes it easier to keep going. For small creators, there are few marketing tools cheaper than this.
The real question is what you share, and how much.
Even in overseas solopreneur communities, building in public is now treated less like a virtue and more like a staged strategy. Early on, openness fuels growth. But once revenue and conversion start to look real, you become easier for copycats to target. On Indie Hackers, people now openly discuss hiding revenue numbers and making product URLs less visible once a business reaches a certain level.
That really means this.
The moment you publish your metrics and revenue, you are showing your growth story to fans while also handing a validated market to someone else. The less capable someone is at coming up with their own ideas, the faster they move when someone else has already proven that money is there. For people chasing revenue with minimum effort, building in public can become less a source of inspiration and more a target list.
If building in public once leaned toward transparency, today it feels closer to selective disclosure.
Share, but hold back the parts that would hurt immediately if copied.
Why I hide important work
That is also why I hide important work from the main blog feed, or only show it in a limited way.
First, important work is not just another portfolio artifact.
It contains how I solve problems, how I read user reactions, in what order I place features, and which message appears at which moment. In that kind of work, the process is often more valuable than the output.
Second, public work becomes AI fodder far too easily.
A few screenshots, a few paragraphs of explanation, and a few lines of user reactions are often enough for someone to assemble a similar landing page and similar messaging in a day. It is no longer that copying takes time. It is closer to: if someone wants to copy, they can.
Third, important work is often better shown to the right person than to everyone.
In hiring, collaboration, proposals, or introductions, a private link is often more accurate. It lets the explanation travel with the work and reduces shallow copying and pointless misunderstanding.
Fourth, I do not put writing and product work on the same level.
Writing is often worth publishing openly. Product work is different. If the experiment is still ongoing, or if it is directly connected to operations, exposure can feel less like visibility and more like leakage.
Copycats probably do not last
The conclusion that keeps resurfacing on Hacker News is similar.
Copycats can replicate quickly, but they rarely sustain a product for long. They want the quick buck, not the long-term attachment to the product itself. When bugs appear, trends shift, or users ask for changes, that is where the gap starts to widen.
AI can generate code and produce a similar-looking interface.
But it cannot truly copy the taste a creator has refined over time, the detail that respects users, the backend optimization that keeps the experience smooth, or the judgment shaped by ongoing maintenance. In the end, what survives is not the code itself, but domain understanding, point of view, and the willingness to keep fixing things.
In the AI era, code alone is no longer much of a moat.
The moat is moving closer to the creator. Not what you built, but why you built it. Not how fast you shipped it, but how long you can stand behind it. That is where the real difference shows up.
So I am not trying to stop building in public.
I just do not plan to lay everything out in the open the way I once might have. I want to share my thinking more openly, while handling important work more carefully. I want to strengthen my own taste, build outcomes that can endure longer, and keep sharpening the standards I rely on.
For now, that feels like the most realistic posture a solo creator can take in the AI era.