The Age of AI Is Not a Classroom. It’s a Jobsite.
Written by Jon Helmberger
That model worked for a reason. It gave us structure. It gave us standards. It helped us build organizations full of smart people who know how to learn, communicate, and solve problems. I’m not saying that foundation no longer matters. It does.
But AI changes that path.
The best way to learn it is not to read about it for six months, take a couple online classes, earn a certificate, and then wait until someone assigns a task. The best way to learn AI is to do something useful with it.
Find an actual problem where you personally understand the context, or sit next to someone who knows the work. Try to solve it together. Ask better questions. Try a different approach. Compare the output of different models. Ask AI what you are not providing or asking. Learn where AI helps, what AI is not good at, where it saves time, and where it creates risk. Then do it again. And again. You are just getting started.
AI is a Craft Built Through Real Work
Nobody becomes useful on a jobsite by watching a video called “Introduction to Construction Excellence,” which I assume exists somewhere. You start with the basic work. Maybe that is carrying material, cleaning up, or learning where not to stand. You learn the rhythm and pace of the crew. You learn what slows everyone down. You learn what quality work looks like from people who have done it for years. Over time, you understand more of the job. You see problems earlier. You know which questions to ask. Eventually, you might be trusted to help run the work, or even bid the work.
AI is similar. It is easy right now to treat it like a title, a strategy, a department, or something close to magic. I do not think any of those views are right. AI is a capability. You build it by using it.
That has been true for me. I have learned more by building than I ever would have learned by watching from the sidelines. I built an educational fishing game for kids that helps youth anglers think through bait type and color based on changing weather conditions. I built an agent to search nationwide for used cars, review new listings daily, and inspect posted photos to verify specific features instead of trusting the written description, because apparently “fully loaded” means different things to different people. I have taken AI-generated document renderings and turned them into editable, usable files. I built a transcription and documentation agent that can turn a speaker’s presentation into a polished, branded playbook. I also wrote a song about my wife, her dog, and her horse that I will respectfully describe as a banger.
None of those started because I needed a training exercise. They started as problems, curiosities, or opportunities. That is the point. You learn differently when there is an actual target outcome or goal on the other side. You either understand the problem better, or you find the gap you need to work on next.
One thing that has become evident to me is how often the first output is just the starting point. The first prompt matters, but the value usually shows up after that. You challenge it. You correct it. You ask it to look at the problem another way. You compare the answer to what you know. You bring your own judgment to it.
That back-and-forth is where the work gets better, and it is also where you get better.
Organizations Need More Than AI Training
I think a lot of organizations are going to miss this part.
They will try to roll out AI the way they have rolled out software in the past. Pick a tool, announce a policy, assign a few champions, schedule some training, and hope productivity shows up like it got the calendar invite. While some improvement might happen through tool introduction and access, the bigger gains will come from people using AI to solve problems inside the team’s actual work.
That means using AI on the work people already recognize… the analysis that takes too long, the report nobody trusts, or simple ideation and refinement. That is where AI becomes useful. Not as a replacement for expertise, but as leverage for people who already understand the work.
With all the noise around AI, I find myself more energized than overwhelmed.
It feels a little like being in my 20s again in the mid-90s, when the internet was suddenly everywhere and nobody fully knew what it was going to become. There was confusion, hype, bad ideas, great ideas, and a lot of people pretending to know more than they did. Eh hem.
So, basically, technology. But if you loved it, it was also fun. It felt like possibility. It was something new to explore that rewarded curiosity, and something that made you want to stay up way too late creating personalized Flash-based animated experiences for the holidays. But I digress . . . .
Why This Matters at SDG
That lines up well with how SDG has always worked.
We have always lived in the space between business problems and technical execution. We are not an organization with solutions looking for problems. We do not stop at the whiteboard. And we will never drop people into seats and call that delivery. Our work requires us to learn the customer’s world, identify whether we are the right company to help, and if we are, create something that provides value.
Our tagline is “Guiding Your AI Future,” and that only works if we are actually experienced guides. You cannot guide from a distance. You have to know the terrain. You have to know where the tools help, where they fall short, where people are likely to get stuck, and what it takes to turn an interesting idea into something useful.
That is why the doing matters so much.
Across SDG, this is not just something our engineers are experimenting with. Our delivery leaders and operational team members are building working software, creating internal tools, testing ideas, and learning alongside our engineers. Our recruiting team has built agents. So has our sales team. So have our business unit leaders and marketing team.
This matters because it changes the kind of guidance we can offer as a team from every corner of our organization. That experience is part of the value customers recieve from us. The more we build, test, challenge and improve with AI ourselves, the better prepared we are to help customers do the same.
At SDG, that fits us. We are builders. We are problem solvers. We are consultants. We learn by getting close to the work, listening carefully, and making things better.
AI does not move us away from that. It puts more weight on it.