Public Apps Are Getting Riskier, So Go Smaller

Ben Thompson, a technology analyst who writes about tech strategy, had been describing a rough stretch for computer security when he offered his prescription for what to build next.

In his words: "The internet as we knew it is dead."

His point is that keeping something open to everyone has become more dangerous and less rewarding. Automated systems can hunt for weaknesses around the clock, and question-answering tools reduce the visits publishers once received.

Much of today's software was built to pass a simple test, whether it worked, not whether it could resist every possible break-in.

Attackers once missed many holes because people had only so many hours to look for them. Thompson says AI removes that limit by testing without pause, so he expects attacks to increase before defenses catch up.

Thompson writes that the public internet will be "a battle zone at best," meaning social media, and "a wasteland at worst" for much else. In that setting, putting a personal site or app on the open web invites trouble.

The alternative he favors is to build for small circles that already trust each other. He points to group chats, which he calls familiar and high trust, as places where people feel safe talking.

His own team offers an example. At his workplace, colleagues have built private tools for podcast production, task tracking, and calendars, none meant for public release. Separately, Thompson says he has built apps for himself to manage his house and sports viewing. Because AI has lowered the cost of making software, a tool useful to one person or one team can now be worth building.