AI Pulse · Under the Radar

There’s a useful reality check emerging around the AI “factory” model. Builders can now generate landing pages, small SaaS products, and marketing assets at almost no cost. The problem is that most of those products still need customers, attention, and a reason to exist.

The critique, echoing a Pieter Levels-style view, is blunt. Some people are running entire factories of AI-generated products, yet almost none have meaningful traffic or revenue. AI has made building cheaper and faster, but it hasn’t solved distribution. It also hasn’t told anyone which customer problem is worth solving in the first place. If a business was unclear on its market before AI, it can now ship that confusion at much higher volume.

For operators, this is the part to take seriously. Don’t measure AI progress by how many ideas, pages, or prototypes your team can produce. Measure whether it improves a real commercial loop, such as finding qualified leads, shortening sales cycles, retaining customers, reducing support load, or helping staff make better decisions. Build less speculative product. Spend more time close to customer demand.

This is the kind of thing we build into an AI command centre, connecting AI activity to the business signals that show whether it’s actually producing value.