> A tool that adds work wouldn't be very useful, would it?

Sure could, if it provides benefits that outweight the additional work. Insurance companies will happily pay a couple more IT-specialists if it reduces the amount of cases they have to cover by an arbitrary percentage. Tools very much can serve other purposes than reducing the friction of human attention.

If you can containerize your infra and thereby mitigate threat vectors that may put your entire enterprise at risk if exploited, that's a good business call to make, even if the cost of your IT-department gets inflated by some margin.

A tool or technology that adds work is extremely useful.

More work means

- more employees needed

- more direct reports for managers

- more jobs for x technology

- more people learning x technology because of those jobs

- more StackOverflow questions for x

- more companies using x technology because it's "easy to hire for"

- more money to be made in teaching x

- more learning material for x

I think the selling point here is getting k8s' benefits using other tools would result in more work than k8s

It reduces work, but imposes a high setup and operating cost. Once you get to the right scale then all the economics look correct. Large teams and organizations and absorb that cost more easily than many who are attempting to use k8s. Of course, being open source, anyone can try k8s, but the mileage is going to vary.

The key question is what are you comparing to k8s in terms of complexity?

Does it add work compared to setting up a VM with Docker and Watchtower? - for sure..

But does it add work compared to setting up something that gives you all the same benefits of k8s without using k8s? - imho definitely not.

Today, we see someone who believes that just because a tool's goal is to reduce complexity, that's what it accomplished