AI agents need new benchmarks

Channel:

IBM Technology

Subscribers:

1,200,000

Published on May 17, 2025 12:00:32 PM ● Video Link: https://www.youtube.com/watch?v=DZp5gX0GW5o

Duration: 0:00

3,112 views

69

Agentic AI is here, but we’re still using chatbot-era benchmarks. Here’s why hybrid, domain-specific evaluations are the future of AI evaluations.