Meta’s Muse Wonder Challenges OpenAI, Why Companies Struggle to Prove AI Returns

Frontier AI model usage is plateauing as cheaper mid-tier models close the gap, while enterprises still can't prove AI revenue ROI.

• CAISI goes White House: Trump administration is moving the top three officials of the AI testing agency CAISI (formerly AI Safety Institute, now 'Center for AI Standards and Innovation') from Commerce into the White House, aligning it with the new SI Task Force led by Jay Clayton under Susie Wiles.
• Muse Wonder beats OpenAI models: Meta's Muse Wonder Full scores 11% above predecessor Muse 1.3 on the Vals Index, trailing only Gemini and some Claude models but surpassing all OpenAI models, per Vals AI's Rayan Krishnan.

Themes, sentiment and reasons are AI-written paraphrases, checked so they never quote an episode. They may contain errors. Not investment advice. Episodes remain the property of their creators.