Back to today’s briefing

Story perspectives

AI Development Report: 43% Fail Model Evaluation

1/16/2025

47 3

Summary

Story summary

Vellum's first-ever State of AI Development Report uncovers a vibrant landscape where over 25% of AI developers have launched enterprise applications, and 39% are in testing. Yet, a staggering 43% neglect model evaluation, risking automation potential. OpenAI shines in model usage, while businesses struggle to pinpoint effective AI applications.

1 of 1

Story summary

Vellum's first-ever State of AI Development Report uncovers a vibrant landscape where over 25% of AI developers have launched enterprise applications, and 39% are in testing. Yet, a staggering 43% neglect model evaluation, risking automation potential. OpenAI shines in model usage, while businesses struggle to pinpoint effective AI applications.