No, AI has not killed creativity in tech
Creativity is alive and well in the tech landscape
It’s easy to look at the tech landscape, especially the role of being a software engineer, and say “Yeah, there’s no longer creativity here. AI has syphoned all of that out.” Or: “Creating software used to be so creative and fun.”
I see this sentiment riddled throughout LinkedIn and X. The road of the nay-sayer is definitely the easiest. Let’s look at the way you can still be creative in tech in 2026.
1. No longer scouring Stack Overflow for a semi-answer
I think there is some nostalgia for this workflow:
Run into a problem outside of your expertise.
Search the web for a problem similar to yours.
Copy the solution to the problem into your code.
Jerry-rig the solution through several iterations until it, by magic, suddenly works.
Feel a rush of serotonin and endorphins for unlocking this puzzle.
The days of this workflow are gone. Because Generative AI can search and solve the problem for us, it may feel less creative.
In fact, just the opposite should be true. Like the little solution from Stack Overflow, the solution that the LLM generates will need to be tested and unit tested. It will need to be jerry-rigged. Accepting it as perfect / non-hallucinated is an immediate red flag for coding in the age of AI.
2. Time to focus on larger, architecture-driven solutions
If you work in a large codebase, you should be able to point your AI-driven IDE at a solution and say, “Do that, except with this.” It will unfurl a large portion of similar code, “except with this” attached. I tend to marvel at these moments of increased productivity.
However, I start to question whether replicating the exact same code with a twist will only bandaid a current systemic issue within our code base.
This is where AI allows us to take a step back and take a look at larger architecture issues. What I hypothetically used AI to create is not very DRY. How could I make this so I don’t repeat myself? Could I create a service? How could I modularize this? Is this performant? How do I check that this way of doing things is performant?
Because I have a better understanding of the codebase, for the most part, than AI, I can take a step back and be creative about improving the codebase besides just being additive.
3. Tools like Cursor unlock new ways to be creative
Cursor rules / commands - I like to think of what I could create to make my coding life even more automated. What is your main source of tedium as a software engineer? What process could you standardize for the rest of your company?
Documentation - I hear ya, documentation is not sexy. But it is creative and important. With Cursor, it knows about a lot of your codebase. What area of the code could use more documentation?
Know an unknown area of the code - Similar to documentation, Cursor is great at asking what a particular area of the code or code snippet is doing. If you don’t understand it, have Cursor point you in the direction of knowledge.
Learn - Similar to knowing an unknown, learning is itself the greatest act of creating. It fires new synapses in your brain. Ask Cursor to go deep into a subject in your codebase you’re unfamiliar with. A Generative AI IDE doesn’t always have to make the answers for you. Use it to learn something new today.
4. Creative outside the scope of a software engineer
Have you been thinking about getting into leadership? What I am freed up from constantly coding, I can explore other avenues of my job that AI currently lacks:
Leading meetings
Mentoring juniors
Giving presentations
Hopefully, you can see how you can still be creative in 2026 in the Age of AI. It takes a little thinking outside the box and little persistence, but in the end, I am a happier dev when I accept that I can use AI as a tool and also maintain a level of creativity at my job.