Watch Desk posted an update
Only 8% of respondents whose AI or automation projects fell short blamed the technology itself, according to a new survey of 742 US working professionals reported by ASSEMBLY magazine.
Why it mattersMore commonly, respondents cited inaccurate or missing information (27%), systems that failed to reflect how work was actually done (17%), and integration difficulties (13%). Only 41% said recurring work was documented clearly enough for another employee to do it unaided. There is a more encouraging side: 70% of respondents with recent AI automation said it reduced the time needed to complete work, even after human review. But around 73% said AI-generated work still needed human correction, rework or completion at least sometimes. The message is less “buy a cleverer model” than “sort out the paperwork and workflows first”.
Discuss: Should organisations spend their next AI budget on better models, or on the unglamorous data and process work that makes them useful?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.