A Dataiku report surveying 685 CIOs reveals that 90 percent believe their careers will be shaped by AI success, while 84 percent struggle to keep up with employee-built agents. In policy and personnel news, OpenAI dismissed three safety researchers for sharing confidential data with a third-party safety group, following reports of ignored safety warnings and the shelving of the GPT-6.1 Astra launch. OpenAI also explored how advanced AI could execute complex scientific tasks by handling coordination and engineering constraints. Meta introduced Muse, an ad-free transaction-based model where merchants pay fees instead of users, leveraging trust to gain a competitive edge. Infrastructure improvements include Lambda's framework for reducing large-scale training compute waste by over 25 percent, and Olmo-core 3, an open framework for scaling mixture-of-experts models to a trillion parameters. Several new open-source decision models, including Clef, Kev, and Decider 2B, were released to help agents programmatically gather context and make decisions. Additionally, Microsoft launched new MAI-Transcribe and MAI-Voice audio models, while researchers debated whether the convenience of AI tools diminishes collaborative human research culture.
Decision models 🤖, Claude-shaped science 🧪, OpenAI safety firings 🚨
Dataiku reports that 90 percent of CIOs feel their careers depend on AI success, though 84 percent struggle to keep up with employee-built agents. Meanwhile, OpenAI dismissed three safety researchers for sharing confidential information, and Meta introduced an ad-free transaction-based model for Muse.
Why it matters
As enterprises grapple with rapid agent deployment and internal governance challenges, shifts in safety protocols and open-source decision models signal evolving technical and operational standards across the industry.