McKinsey predicted that there'd be 900,000 cell phones by 2000. If their automation prediction's error will be even a quarter as off as their cell phones', then

McKinsey predicted that there'd be 900,000 cell phones by 2000. If their automation prediction's error will be even a quarter as off as their cell phones', then we are not going to have enough people to have their jobs automated. I know inference here is inductively fallacious and here we're talking about job categories not job numbers, but it just makes 49% feel like 98%.