A glance into the past as I accelerate into the future

I started learning programming back in 2016 because I wanted to build an app and I only knew one programmer. He was, of course, not interested in building my app. Everyone else had already asked him to build an app. He was sick and tired of it. So I figured I had to do it on my own. How hard could it be, right?

After banging my head against the wall, quitting a few times, taking a few coding classes at the university in Bergen and landing a gig as a junior developer in Sbanken, I finally cracked the code. Pun intended.

During the day I hammered the keyboard on my windows desktop in an attempt to soak up as much knowledge as possible from the senior developers at work. At night I built games in Swift, and got my first game, The React Attack, published on the app store. It got downloaded 200 times and amassed $1 in ad revenue. I felt alive.

I also built huge projects in .NET. Both alone and with friends who also aspired to build the next big thing. Of course, none of them ever launched. Mainly because I/We got overwhelmed, tired of the project or came up with an even better idea. Back then we had no coding agents, no Vercel, no Supabase or Next.js, deploying apps was a hassle for a newbie. And getting stuck was a huge pain in the ass. I spent more time than I care to acknowledge crawling posts at Stack Overflow about obscure runtime errors and bugs I came across.

At the end of 2022, I, as everyone else, got introduced to chatGPT and it didn't take long before I copy pasted code back and forth between chatGPT and VS code. My velocity increased tenfolds. The quality? let's not get into that.

Now, in 2026, I have agents so advanced I rarely need to write a single line of code in my projects. I just direct. The app I spent weeks building back in 2018-2019 can now be built in less than 30 minutes with Claude code or Codex. Most likely even faster than that. It baffles, terrify and excite me at the same time.

It's exciting, because I now can try to solve challenges I face in product management in new and different ways. I'm not saying AI can or will solve product management. But I think it makes it easier to create tools that are more dynamic and context aware. And it's so much cheaper than it ever has been before ever. Which makes it worth keeping an eye on as we jet towards the future ahead.