Don't be a generalist, be a connector. It's not the same thing.

For a long time, extreme specialization has been a reliable way of being valuable in the marketplace of skills. From how salaries for niche skills like Cobol or Blockchain are significant above the average, to that senior developer colleague of yours who’s the only one who knows how to deploy and debug those old C application your company still uses, you could build a career on being one of these hyper specialized types.

I call this value through obscurity: knowing something few know, especially if that something is not well documented or widely available. I don’t mean it in a bad way, but Cobol developers have secrets. They know things about those mainframes that are not readily available at just a quick search. Javascript developers have no secrets. Everything you could ever want to know about Javascript is documented all over the internet.

Yes. It has been changing for as long as documentation has become better and more easily accessible, and now with AI it’s changing faster than ever. Suddenly, that colleague of yours who’s the only one who knows how to deploy and debug those old C applications doesn’t seem so irreplaceable anymore. Sure, you’re not going to be able to be as good as him just by using AI, but at least you can try, and you might just be good enough. The value of obscurity is decreasing, as more and more skillsets are brought into the light and borged up by the models, to be then spewed back up on demand for the low low price of 20$/month.

Look, I’m not saying this will disappear over night, but it’s under attack.

If hyper specialization is not reliable anymore, should you then become a generalist? Know a little bit of this, know a little bit of that, do a little bit of this, do a little bit of that?

Yes … and no. There are two types of generalists. One type has debatable value. The other one is priceless.

This is Bill. Bill is the manager from the movie Office Space. Bill has no vision. Bill has no ideas about how to improve the company’s products. Bill has no meaningful relationships with key people. Bill doesn’t motivate or inspire. Bill has no technical knowledge in any area. Bill is generic butter spread too thin. He could be replaced by a well trained German Shepherd, and it would be a massive improvement.

But ok, Bill is a manager, managers can be like that. What about technical people?

This is Fred. He’s as generic as the AI image I generated to represent him. It’s not his fault, he might actually be a great guy if you got to know him. But you won’t. It used to be that being Fred was good enough. Fred used to be the middle class back when things were simpler. Fred once in a while asks some decent questions at the planning meeting. He usually does the same kind of task. What he does is not unique, but he’s good enough. He's not really interested in what happens upstream or downstream from his team. Once every couple of weeks he suggests something a little new, but almost never really impactful.

More and more, being Fred is not going to be enough anymore. Don’t come at me, I don’t make the rules.

This is Don. Don’t is a visionary leader. Don is known and valued by customers. Don is known in the industry. More than that, Don is technically excellent. He’s not the best at every detail, but he has the best overall understanding of what great work in his industry means. Don makes bold decisions. Experiments. Cuts ties when he needs to and forges new ones. Takes chances. Don is a movie character playing a 60’s ad executive so the particulars are irrelevant to us, but in principle, Don is exactly what I’m talking about. Let’s look at a closer example.

This is Mike. Mike (not real name) is someone I know in real life. I didn’t ask him for permission, so his face is not here.

Mike is a software developer. At a glance, Mike is a generalist, he’s not hyper specializez in anything. But a week doesn’t pass without Mike spending a few hours diving into something new: a new tech, an obscure part of the project, a possible alternative solution. Curious and open minded, Mike is always tinkering with something. Mike is also interested in what the other disciplines do. He is close to the BA and QA. With some regularity, Mikes likes to meet with clients, see how they think, what worries them. Sometimes this happens naturally, sometimes he volunteers for the opportunity. Mike is in touch with other teams from the organization. When Mike hears the requirements, he can intelligently propose alternatives and improvements that makes sense, for the clients, the business, and the overall architecture. Sometimes he evens comes up with his own ideas for new features, and sometimes the clients love them. When there’s a bottleneck, or a friction between teams, Mike has the relationships to figure it out and reach a compromise, even if he’s not formally a manager. Mike is pretty well known in the IT community in his town. He’s admired and appreciated for his contributions to communities and conferences. He’s seen as a good teacher, a motivator and a multiplier of a team’s productivity.

Mike is the kind of generalist you want to be. The priceless connector who combines breath with depth, curiosity with focus, technical excellence with human skills. You want to be like Mike.