Sources
Greg Brockman: OpenAI Staff Hate It When a Coworker's ChatGPT Slacks Them — Even for Work They'd Happily Do if Asked Directly
Willison collected an August 1 quote from OpenAI president Greg Brockman: 'At OpenAI, many people hook their ChatGPT up to Slack. People really don't like when a coworker's ChatGPT contacts them asking for help with a task, even when they'd be perfectly happy doing that same work if asked by that coworker.' It is a rare piece of first-party evidence about agent-to-human social friction from inside the lab shipping the agents, and it isolates the variable cleanly — the task is identical, only the requester changed. For anyone building agents that initiate contact with humans, this argues the failure mode is social rather than technical, and no amount of capability improvement fixes it.
↳ Follow the thread