I'll give you an example. The entire justification for the ROI on three different systems I have built in the last 5 years was that they wouldn't have to hire the dozens of people required to do the job without automation. In my current job, we are currently looking for ways to automate currently manual processes for purposes of reliability, but they also have a side effect of reducing labor costs.

I have been personally responsible for automating dozens of jobs throughout my career. These weren't layoffs in most cases - these were people that would have been hired without this software.

Paralegals are becoming more rare precisely because lawyers can search more effectively, are they not? Due to the search , one lawyer can be more effective than two decades ago, no?

One secretary services multiple Vice Presidents and directors in many organizations, and handles extra duties on top of that.

It isn't eliminating all jobs - it's making one person as efficient as three, five, ten, and even more were a generation ago.

By reducing the cost of some things sometimes it is offset by using more of it. Sometimes you unlock a lot of latent demand. Ford created jobs for autoworkers despite increasing productivity by 10X. Uber is creating jobs by lower the cost of a taxi ride.

Word perfect revolutionized document production at law firms. Before you had to actually take a red pen to a document and write in changes. If someone else gave you a document, you'd have a secretary go word by word and underline changes. Now that stuff can be done in 3 seconds with a plugin in word.

So what happened? People exchange drafts like 10X as often because it is easier. Sure some sectaries have been lost. But it's increased the need for lawyers because more work is done that couldn't before. And laypeople produce probably 50x more documents than before email and word processing.

Legal research: before you'd look it up in a book that contained key cases. You'd find something or not. Now with Lexis and Westlaw giving you way more cases to look at it, I wouldn't be surprised if we end up researching more. People dig for a needle in a haystack.

And if you want to see how far we've got to go, legal research is still mostly done by having a human read the case and create keywords and classifications. It's still more accurate that Google is for legal issues. But I bet that won't last long. Frankly I'm shocked searching isn't better.

But most professionals aren't doing repetitive tasks a machine is better at. For lawyers, the only real easily picked fruit I see is document review. People are trying what we call "predictive coding" now. But those jobs were already outsourced to not-real lawyers who worked temp jobs.

I sure wouldn't want to be a data entry person or a secretary. I really really wouldn't want to be a worker in a factory in the third world.

But we are far away from software eating the professional world. Software has already taken the easy gains away from these types of jobs. The accountant isn't spending a lot of time crunching numbers.

To increase productivity by multiples now, you'll need artificial general intelligence. If you can do that, we are replaced anyway, I hope our corporate overlords are merciful.