a16z discusses TypeSafe AI’s Jev, a model that returns choices with confidence

a16z published a conversation with TypeSafe AI’s Diogo Almeida, Ben Horowitz, and Martin Casado about Jev, described as a model built to live inside software.

Almeida’s elevator pitch asks where the automation is. AI is very capable in chatbots and coding agents, but software cannot readily use text written for people to read. Jev reads natural language and returns a choice from a set of options, assigning a confidence level to each, so developers can build programs that reason about intent and make probabilistic decisions without relying on a person to interpret the output.

TypeSafe AI’s stated philosophy is “We build prod, not God.” The conversation is available as a YouTube video.