decision model

A language modellanguage model, n.: A system that has learned how likely different sequences of words are, so it can predict the next word in a text and, by doing that repeatedly, write text of its own. that takes unstructured text but returns typed numerical outputs instead of generated text: a probability for a yes/no question, a probability distribution over a fixed set of choices, or a value on a numeric scale. According to TypeSafe AI, several questions about the same input can be answered in parallel at roughly the cost of one, and since nothing is generated, pricing can rest on input tokenstoken, n.: The small chunk of text an AI language model reads and writes — often a word, part of a word or a punctuation mark. AI usage, limits and prices are counted in tokens. alone. Decision models target work such as spam detection, labeling, prioritization and ranking, where a chat modelchat model, n.: A language model trained to hold a back-and-forth conversation, keeping track of who said what and replying as a helpful assistant. It is the engine inside chat apps.'s prose would otherwise have to be parsed into a decision. They revive the output format of a classic classifiertext classification, n.: Sorting pieces of text into categories automatically — what topic an article covers, whether a message is spam, which team should handle a customer's complaint. but, unlike a classifier trained for one label set, accept new questions at inference time. Critics note that a bare score cannot explain which part of the input drove it, which makes careful evaluationevaluation, n.: Testing an AI system to find out what it can and cannot do, how well, and how safely — from exam-style quizzes to hands-on trials and expert reviews. and bias testing essential.

We swapped the chat model for a decision model and our triage pipeline stopped choking on malformed JSON.