The Asana app already has workflow automation instruments, however the addition of LLMs permits for extra highly effective automations. “What’s actually thrilling about embedding LLMs into work administration is that there’s judgement, there’s nuance,” stated Costello. “As a substitute of it being a black and white rule — ‘if these items are lacking, ask this’ — it’s directional, like, ‘right here’s what we sometimes search for; use your judgment.’”
That stated, given the unreliability of LLMs, there’s potential for the teammates to introduce errors into the work administration course of.
Costello stated it’s essential to have a “human within the loop,” when the AI assistant takes motion. “We need to be sure that people are the last word deciders,” she stated. “They’re those which might be accountable, they’re those finishing the work: AI is simply serving to. AI shouldn’t be transferring work ahead in a manner that’s not permissioned successfully by the workforce or the individual driving that work.”
