Editorials

Do You Consider Robots (Automation) a Threat?

Increasingly having conversations (including one on FB in the last couple of days) about automation and robots and what it means to jobs and duties… so I’m curious what you think of it all?

To me, it seems pretty likely (unavoidable?) that things that can be decided based on data will some of the most prime – and learning from masses of information (here’s the call to big data) also lends itself to this.

Just in the last several days there are articles that suggest that robot doctors will be able to use Google’s AI to diagnose cancer more quickly than human doctors. And at the same time, we’re hearing more and more about self-driving cars, about automation of systems…

To bring this all home to our work specifically, today, in many of our projects, there is already substantial automation. If a server goes out, it replaces itself, complete with code updates, testing and then sending an email to us when it’s done.

If systems need additional capacity, same thing. The automation steps in and scales out the servers and operations, then shoots a quick email as it scales up, and scales down when the heavier loading is relaxed.

These are pretty basic, but it shows that things that rely on factual happenings are already there. Azure self-tunes databases and operations and performance. Many cloud providers will optimize servers automatically, will even shut down run-away processes. I’ve written here before, too, about the Azure and its ability to “sniff out” transactions and activity that seem suspect, including making suggestions to you on what needs attention.

One of the things that always bugged me when writing code (I was a Pascal guy), was “missing ;” – well, if you know where (you just gave me the position and line number mr. error message), WHY NOT JUST ADD THE DUMB THING?!? Yep.

Clearly the automation is coming forward on administration, troubleshooting, configuration, performance, etc. I welcome it. I think it will be faster to recognize issues, make faster decisions, etc.

Of course all the danger areas are still there. Terminator movies, anyone? Sure, those are a bit away (I hope) but you can pretty quickly see that decision point for automation where they are picking what stays, what goes… WHO stays, WHO goes.

Do you consider this automation a threat, or a leverage point for your work with systems? Do you think it’s something that will enhance the employability of an administrator if they know how to work with automation solutions and things that are screaming ahead?

I, for one, welcome it. At the same time, I think the impact of full-on automation will be stunning, to say the least. I believe it’s a planning horizon thing, and something that warrants attention and thought and management now, not later. Not in a conspiracy type of way, but in a not sticking our heads in the sand and pretending it’s not arriving sooner than we think, kind of way.

Facebooktwittergoogle_plusredditpinterestlinkedinmail