Don’t Be Scared of Automation. Be Scared of Someone Using It.

Dark title card for a blog post: Don't be scared of automation. Subtitle: Be scared of someone using it. Chips read Operator, Always Be Learning, and Human Problem.

TL;DR

Automation isn’t going to take your job. The person who has decided to operate it might. The next challenge in network engineering is human, not technical: helping teams adopt automation safely, and knowing the difference between when AI is right and when it is confidently wrong.

I share my personal work publicly because I’ve found it to be a method for creating and maintaining a “human” networking and opportunity pipeline. In my experience, you never know when you’ll need to tap into that pipeline’s information stream. In this modern technology world, though, having that pipeline on tap is absolutely a strategic advantage. It also demonstrates and reinforces one of my personal tenets: Always Be Learning (ABL).

Sharing the work does two things at once. It puts you in the stream with other people who care about the same problems, so the next idea, the next conversation, or the next opportunity is easier to find when you need it. It also keeps you honest. If you’re going to write something down and put your name on it, you have to keep learning it well enough to stand behind it. That’s ABL in practice, not as a poster on the wall.

That matters now more than ever.

Don’t be scared of automation. It isn’t going to take your job. But do be scared of someone using automation. That’s the threat: people who have accepted that automation will exist, and who have decided to be the operator of the automation rather than let it run them over.

You can take that as a warning, and it is one. You can also take it as an invitation. Automation is already in our networks, our changes, our tickets and our security tools. Hoping it stays somewhere else is not a plan. The safer path is to become the person who can operate it: to put it to work on purpose, to know what good looks like, and to stop it when it doesn’t.

The Next Challenge Isn’t Technical. It’s Human.

The challenge I most want to solve next in network engineering isn’t technical at all. It’s human. I want to help teams and organizations learn how to embrace automation and integrate it into their networks in ways that are safe, secure, efficient and productive. I want to demonstrate responsible and effective network automation that has a substantial and positive impact on security posture, and a real and meaningful ROI for the C-suite.

That is not, first, a shopping problem. Most teams can find a tool. The harder work is getting people and process to change with it. Someone still has to decide what the automation is allowed to touch, who is watching it, how you tell whether it actually improved your security posture, and how you show the business that the investment was worth it. If those questions go unanswered, you don’t have automation with a meaningful return. You have a tool nobody is really operating.

Safe means you have thought about what happens when the automation is wrong, not only when it works. Secure means it doesn’t quietly become a new way in, or a new way to make a change nobody can explain later. Efficient and productive mean it actually saves time and reduces toil, instead of adding a second system that people work around. And if you want the C-suite to care, the result has to show up as a stronger security posture and a return they can recognize, not just a busier dashboard.

I’ve been writing about the safety side of this on the site, including enterprise AI guardrails for blast radius and keeping humans on the loop, and about how agents should connect and be observed in the Agent Interconnection Model (AIM). Both exist because the wiring only helps if a team can run it responsibly.

We see it playing out in real time. The bad guys are using AI at machine speed to hurt our companies. The good guys have the same chance to use the same tools to defend against them. The challenge is that the good guys aren’t always aware of how to use AI in that defense, or how to structure themselves around automation so the defense actually works. Access to the tools is not the scarce part anymore. The judgment is, and so is the operating model: who does what, what the automation is allowed to do, and how the team stays in command of it.

That’s the area I want to help with. I want to help the good guys get comfortable enough with automation to use it on purpose, and disciplined enough not to hand it the keys and look away.

The Special Sauce Is Becoming a Commodity

I believe that’s the next level for network engineering, looking through a broader industry lens. The “special sauce” in technology hardware, and even in technology silos, is becoming a commodity.

That used to be a harder thing to say out loud. For a long time, knowing a particular platform, a particular vendor’s way of doing things, or a particular silo (network, systems, security, voice) was enough to set you apart. That knowledge still matters. I’m not saying it doesn’t. You still have to understand the network. But everyone has reasonable access to the same technical knowledge now. The docs are public, the labs are easier to stand up, and if you have a question, you can get an answer in seconds. Sometimes that answer is right. Sometimes it is confidently wrong.

That access used to be the gate between employed and unemployed in the technology industry. If you had the knowledge, you had a path. If you didn’t, the door was harder to open. The game has changed, and it changed quicker than a lot of people realized.

If the same knowledge is widely available, the differentiator has to move. It moves to whether you can apply it, whether you can tell good output from bad, and whether you can use these tools to make better decisions and more reliable changes, not just faster ones. Faster and wrong is not a promotion. It’s a bigger blast radius.

The New Gate

The gate between employed and unemployed in technology now is:

  • Do you know the difference between when AI and automation do it right, and when they are confidently wrong?
  • Have you figured out how to use AI to magnify your strengths without also magnifying your weaknesses?
  • Have you figured out how to use AI to allow you to make more confident business decisions more quickly?
  • Have you figured out how to use AI to allow you to execute more reliably?

Those are the challenges I’m looking to solve next. Always Be Learning still applies. What we have to keep learning has just gotten bigger than the next command or the next chassis. The people who will do well are the ones who decide to be the operators of the automation, not the ones who hope it stays out of their way.

Found this useful?

Everything here is free and stays that way. If something I wrote saved you a headache, you can support the site on Ko-fi.

Support on Ko-fi

No responses yet

Leave a Reply

Your email address will not be published. Required fields are marked *

Latest Comments

No comments to show.