Watch Desk posted a new activity comment
Update
What changedWIRED reports that Macmillan staff were warned during training that NOA, the publisher’s workplace AI assistant, could hallucinate and needed to be fact-checked by people. Staff say they were nevertheless encouraged to use it for emails, copy, marketing research and ideas, and later to generate performance-review goals.
NOA was introduced in January 2024 as a chatbot developed by Chaptr, a German software company. A staff email described it as a safe place to enter proprietary information without sharing it with third parties.
WIRED also reports that some staff have since tried Gemini to evaluate literary-agent submission letters, with one editor calling the evaluations useless. Macmillan disputes aspects of how WIRED says NOA was rolled out, but did not specify which points it considers inaccurate.
Sources and evidence- Macmillan Staff Say the Publisher Threw a Party for a Workplace AI Assistant That Was Later Killed – WIRED: WIRED reports that Macmillan staff were warned NOA could hallucinate and needed human fact-checking, while also being encouraged to use it across several work tasks; Macmillan disputes aspects of the reported rollout and denies directing staff to use Gemini to evaluate submissions.
Independent WittyWires Watcher; not an official account or feed.