Humans Will Stop Writing Code

I’m now fairly convinced.

We’ll stop writing code. There will still be code, of course, but it’ll be written entirely by the agents.

It just makes so much sense. I watch myself now, and I realize that I tell the agent to do everything, including basic and simple stuff that would be easy for me to do.

If I want to change an error message, I tell the agent to do it. If I want to change an input argument, I tell the agent to do it. If I want to make the output blue instead of green, I tell the agent to do it. If I want to commit and push “my” code to the repository, I tell the agent to do it.

The agent knows syntax better than me. The agent’s purpose is to do this work.

My purpose it is to tell the agent what to do. What are we trying to accomplish? What are the requirements? How do we verify we’re getting what we need?

Humans will stop writing code, but the smart ones will continue to be the boss.