There's enough sublety and variance in benefits (to employee, employer, and/or customers) of "a job" that I don't think there's any simple answer. In most cases, the benefits are not evenly distributed, so you hit the problem of interpersonal comparison - if you hurt 100 and help 1000, by hard-to-measure amounts, is that good?

My ideal definition of "job" is "opportunity to perform/produce something of value to others, and to capture enough of that incremental value that you're better off doing it than the next-best thing you could do with that time/effort".

It's clearly better for the company's owners, the company's customers, and the company's vendors to automate: things are faster and more accurate without manual data re-entry. Some of the 100 laid-off employees will benefit by the realization that there's easier or more rewarding work they can do (improving the automation, or harder-to-automate data entry somewhere else). Some of them will be harmed because that opportunity no longer exists, and their next-best opportunity is not as rewarding.

Is it overall beneficial? Depends on your aggregation function. But it doesn't really matter - it's enough improvement to customers and vendors that the company needs to automate to stay in business at all.