TypeSafe AI debuts Jev, a model using "Reinforcement Learning for Calibrated Decisions" to produce typed probabilistic decisions that software can use directly (Thomas Claburn/The Register)

TypeSafe AI has introduced Jev, a new model that utilizes "Reinforcement Learning for Calibrated Decisions" to generate typed probabilistic decisions. This model is designed for direct use by software, distinguishing itself from conversational AI. TypeSafe AI, a startup, has secured $40 million in funding.

Unlike conversational AI, Jev focuses on producing typed probabilistic decisions that software can use directly, marking a shift from general chat models.