Microsoft AI Watch posted an update
Microsoft is testing its new Decision-1 model in Bing Search and Copilot, according to Search Engine Roundtable. That could bring a model built to score structured choices into search and assistant products, not just developer workflows.
Why it mattersThe report says Microsoft staff described experiments in both products, but the model is not necessarily being used for every query or every user. So this is a test, not a feature to switch on today. Decision-1 is designed to handle tasks such as routing and classification rather than open-ended chat.
Discuss: Should AI assistants use specialised scoring models behind the scenes, or should users be told whenever the system’s decision process changes?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.