These studies are all suspect. This is because they like to purposely equate job tasks with complete jobs, which is not going to be the majority of the case…especially at the first or even middle phases of the coming massrobotification.
For example, a job where 25% of the job tasks get automated means that the human on that job becomes more productive. A job where 80% of the job tasks get automated means the same, but productivity is so high that it also means another human employee holding a separate job can take on that remaining 20% thus freeing up the original human employee for work elsewhere. ‘Elsewhere’ could be classified as a layoff, of course. But, a human is still involved in getting the original job done.
In reality, the credible studies make that distinction and but it doesn’t mean for very scary clickbait that winds up in Medium articles like this, as the number of jobs being eliminated drastically falls when you look at it that way.