The number of mobile apps on Apple's App Store rose 30 percent last year, yet downloads grew only 3 percent. Artificial intelligence now writes code from plain language, flooding markets with software — but where are the next Slack or Candy Crush? A. I. hasn't killed the "killer app"; instead, a different digital revolution is emerging.

People resist switching software unless necessary, and A. I.'s real strength lies in reducing small frictions rather than replacing core tools. An M.

I. T. study found up to 95 percent of corporate A. I. productivity efforts fail.

Big A. I. companies rush model releases without manuals or clear tutorials, expecting users to "prompt harder" when projects stall. The tech industry, despite its disruptive image, is surprisingly set in its ways.

Venture capital pours almost exclusively into A. I.-first products, making traditional apps seem risky. Since A.

I. makes software cheap and cloneable, code loses its "moat" and becomes a commodity. A. I. excels at reproducing familiar patterns — like a flimsy PowerPoint clone — but struggles to create genuinely novel experiences.

Its true value may be unlocking previously inaccessible computing capabilities, such as complex business models or browser-based machine learning systems, rather than generating the next viral app.