Skip to content
AI IntelligenceJul 30, 2026AI Intelligence
Article

A former OpenAI researcher predicts that $100 billion will flow into training data because scaling alone won't be enough.

He argues that models are becoming more specialized in coding and math while stagnating or regressing in other areas.

Data Cube AI EditorialSource: The Decoder
01

Source Brief

A former OpenAI researcher predicts that $100 billion will flow into training data because scaling alone won't be enough. He argues that models are becoming more specialized in coding and math while stagnating or regressing in other areas.