Challenge Is Gone · Listed Articles

Alex Gusev, CTO of Uploadcare, reflects on turning 40 and realizing that AI coding agents have removed the engineering challenges that drove a lifetime of side projects and startups, and on what it means to look for a new kind of challenge in a second professional life.

All my life I have consciously and unconsciously strived for different things. Consciously, I pursued the goals society set for me: make money, be useful, have children, and so on.

My unconscious goal was always the same: to challenge myself and try to succeed. It might seem that my conscious and unconscious goals complement each other perfectly, and in part that's true. The only question is which challenges I chose for myself.

I spent a lot of resources (money, time, and possibly health) building my own businesses, none of which became successful. Only by the age of 35 did I begin to understand why. (Yes, I'm not a very fast thinker.)

I was only ever interested in engineering challenges. Could I build an app that finds optimal public transport routes (back when neither Google nor Yandex had one)? Could I release a game on the App Store? Could I build my own service for booking car wash appointments (by the way, I still need one)? Could I build a service that translates books into different languages?

Due to some quirk of my character, I always saw things through to the end. And in the terms of my unconscious, the end is the state where the product is available for anyone to use. After that I instantly lost interest, because the problem had already been solved. Make my product popular? Make money from it? Completely uninteresting, because that's not part of the challenge.

I just needed to prove something to myself.

This year I turned 40, and the challenge is gone.

I simply know that I can solve any engineering problem (that I can come up with) with the help of artificial intelligence. I've tried it a few times — it works.

Just a year ago, I spent several months cleaning up what I had vibe-coded for https://lettercarta.com. Now I sometimes don't even look at the code the agents write, because I've gotten used to the fact that they will simply do certain tasks well. And if they make mistakes, other agents will find them and tell me.

I used to choose my challenges by weighing my own abilities, because it was extremely important to me that they be hard but achievable. I had to avoid disappointing myself and still experience real joy. An easy stroll doesn't bring the kind of joy an adventure full of hardships does.

But now there are no more challenges.

It might seem that because of this I've lost my purpose as a professional and don't know what to do now.

In fact, I'm amazed at my luck. Some folks built large language models, then multimodal models, agents and everything that comes with them, and freed me from that challenge->achievement loop, which was always incomplete.

Now I'm completely absorbed in thinking about what to do rather than whether I can do it. I see how many different products people build, and I don't want to build yet another one, because I know I'll pull it off.

My whole life there was a staircase in front of me, and I kept climbing onto the first step and then stepping back down. Now I've climbed onto that step, and I won't come down from it.

Now I have to learn to think all over again, because it's impossible to solve a problem I can't set for myself. And the number of problems I can solve without knowing it yet has grown enormously in recent years.

A couple of recent examples:

I found a way to create an image that is a memory bomb for a certain image processing library. The image is less than a kilobyte, but open it with that library and it takes up 10 gigabytes of memory. A colleague shared a suspicion about this, and I suddenly asked myself: are problems like this now in the set of problems I can solve?
I have a scanner that works on Windows but doesn't work on macOS, even with the manufacturer's official drivers. Or rather, didn't work. Now it does, and as a bonus I have the most convenient image scanning app I've ever used. It took me an hour and a half.

Why is this luck? Because I personally did nothing to earn the chance to live two lives in one.

In my first life, I was sure it was impossible to create intelligence on a von Neumann machine. My second life became available to me because someone built something better than intelligence on a von Neumann machine. (Yes, I'm still skeptical about calling it intelligence, but that doesn't matter.)

I could have died as someone who never climbed onto the second step. And now I have a chance.

P.S. I am a human and can make mistakes. The spelling and punctuation in this post were corrected with AI.