OpenAI’s Jev clone could help the frontier lab stop its swarming agents
techcrunch ·
During OpenAI’s Dev Day, CEO Sam Altman introduced the Decisions API, a new tool designed to help manage AI agents by providing fast, low-cost decision-making capabilities. This product mirrors Jev, a model released earlier by TypeSafe AI that functions as a high-speed classifier for software automation tasks. Altman explained that the API allows OpenAI's Luna model to choose from predefined options efficiently while maintaining safety and language support features. The move highlights a growing industry trend where developers seek cheaper and faster alternatives to large language models for specific automated decisions. TypeSafe AI’s CEO Diogo Almeida noted that this interest signals a shift toward 'System One' style intuitive thinking in AI architecture. These specialized models are particularly valuable for monitoring AI agents, offering a cost-effective way to detect and block malicious actions compared to using frontier LLMs. For instance, monitoring costs can drop significantly when using these lightweight classifiers instead of expensive general-purpose models. As more tech giants enter this space, the focus remains on balancing speed, cost, and the intelligence of the outputs.
techcrunch