
Specialized apps may shape the AI era more than tech titans, useful context for a colleague tracking innovation trends.

Apps Will Win the AI Boom Story flow and key facts
Venture capitalist Chi-Hua Chien, known for early bets like Facebook, predicts that the largest financial gains in the AI era will come not from infrastructure companies like NVIDIA or OpenAI, but from consumer-facing applications that use AI to solve everyday problems. Drawing on 20 years of tech cycles, he notes that in past revolutions—PCs, the internet, mobile—the most value accrued to software apps, not the underlying platforms. During the mobile era, infrastructure created $700 billion in value, while apps like Netflix, Spotify, Meta, and Uber reached $3.7 trillion in market capitalization. Chien believes AI models will soon become commoditized, shrinking the performance gap between high-end systems and smartphone-based models from two years to just three months by next year. This shift opens opportunities for startups worldwide, especially in healthcare, fintech, entertainment, and personal digital assistants. He also warns that vertically integrated venture firms are inflating valuations through rapid financing, making competition tougher. Ultimately, Chien doubts a single 'super-app' will dominate, favoring specialized services that deeply understand human behavior.
Facts
- VC veteran Chi-Hua Chien predicts AI applications—not infrastructure companies—will generate the most value in the AI era.
- During the mobile era, apps created $3.7 trillion in market cap versus $700 billion from infrastructure.
- AI model performance gap between top systems and smartphones is expected to shrink from two years to three months by 2027.
- Chien identifies healthcare, fintech, entertainment, and digital assistants as the most promising AI application areas.
- He warns that vertically integrated venture firms are inflating startup valuations through rapid financing rounds.
Canto visual news explainer. AI tools may assist production. Editorial policy





