AI model evaluations

What experts have said about AI model evaluations

5 statements · 2 positive · 3 negative

1. Nathan HagerPositiveSep 21, 2026· Stock Movers

   Accenture will evaluate models, test safeguards, and assess alignment with human preferences.

   “So Accenture will be used to evaluate the newest models, test safeguards, and ensure the technology is aligned with what humans actually want.”

   Listen at 3:43

   Open the episode · Novo Declines; Critical Metals Soars; WBD Higher
2. Noam BrownNegativeSep 17, 2026· Dwarkesh Podcast

   Short release cycles may prevent full evaluation of long-horizon AI capabilities.

   “If you're in a world where they can operate effectively over 3 months, but the model release cycle is every 2 months, then you don't have a way to evaluate the models at the full length of their capabilities before the model release cycle”

   Listen at 1:03:57

   Open the episode · Noam Brown – Agent swarms, alignment, & recursive self-improvement
3. Noam BrownPositiveSep 17, 2026· Dwarkesh Podcast

   Realistic evaluation environments could help predict real-world AI behavior.

   “if you can create very realistic environments and put the AIs in there. If you have a sufficiently realistic evaluation environment, then you can get a sense of, okay, is the AI actually going to behave well when we deploy it in the real world?”

   Listen at 1:16:02

   Open the episode · Noam Brown – Agent swarms, alignment, & recursive self-improvement

Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.