There was a chart drawn for me that I’ve also drawn for many folks over the years. It looks something like this:
Career Impact Over Time
This is someone’s impact over time in their career working as an individual contributor. Somewhat linear growth as they learn and develop skills with some variation along the way.
But at a certain moment in time (the orange line) an engineer will have reached the maximum amount of work they can do in a day, month, year, etc. This can take years or decades depending on the person and their job.
This peak notes the maximum they can achieve by themselves, and impact can only grow higher by influencing others. This can be a turning point for some engineers, some deciding to turn to management or a deeper introspection on what they are doing with their life/career/project. Some find leadership and mentoring and end up truly enjoying it without being a manager.
Types of Engineers
There are many types of engineers that can struggle to break beyond that peak, and honestly there isn’t anything wrong with that, a great engineer is a great engineer. Oftentimes folks in that space have a strong desire to continue to improve their impact or in some cases just make more money or have a better title.
2 that stick out in my mind are folks with a big ego who have to be seen as being the hero, or folks that are extreme introverts that want to do whatever they can to not interact with others. I think both of these types of folks just haven’t found the right way that works for them and everyone has that moment where they have to figure out how to do it that isn’t opening an editor.
AI Agents
The interesting thing for me here is the emergence of AI agents or agentic coding. This type of coding is essentially having an AI bot write code, or solve problems, for you. You can have several agents running in parallel across many different facets of a project from writing code to improving security to adding unit or integration tests. The code these agents can write can have varying quality, and some of it or maybe a large portion of it is dangerous to use in production without some sort of review.
For engineers, particularly the 2 mentioned above, will these AI agents give them another mechanism to increase their impact and produce more work than they normally could in a 24 hour window? AI agents can work through the night.
I don’t think this completely eliminates the need for engineers to work with other people. Half the problems of software engineering are just working with people and not in the code. But does this give an outlet for people to produce more and have a higher impact by managing AI bots and reviewing the work at a senior level? Or does this actually cause more issues because the amount of domain space to be covered in producing these? Can AI help these engineers stuck at the peak to break through this barrier?
What even will be a principal or distinguished engineer in the next 5 years? I simply don’t know.
Naming will remain the undefeated king of troubles in software engineering.
Pete Shima 