Agentic coding's hidden productivity boost
Why exactly do coding agents make us feel so much more productive as developers? I think Max Hodak hit the nail on the head at end of a recent Naval pod:
The thing that really changed is—it used to be you could build a lot, a lot would go straightforward, but then you’d hit some random thing and you could spend an indefinite period of time debugging some narrow thing. Now, with agents, you just don’t get stuck anymore. Which is pretty amazing. Relatively quickly they can find the right way to do things. It used to be that—I remember when other friends would try to learn to program, it was like—”Nope, it’s intrinsically frustrating. That’s part of the deal. That’s how you learn.”
And that just isn’t true anymore.
One thousand percent this. And good riddance! Back when I taught beginning web development with Ruby on Rails, I used to describe my job as 10% teaching and 90% unsticking people. You know, junk like:
“Why can’t it connect to Postgres?”
“For no reason the cookie is now empty?!”
“Why does ActiveRecord not return any records?”
I’d answer umpteen of these questions on the daily to ensure people could make progress on what they actually wanted to accomplish. And how did I know the answers? Because I’m stubborn, that’s how. Through pain and suffering, I learned those particular fixes to those particular problems at that particular time. But that hard earned knowledge became stale and mostly useless within a few years, max. Not anymore, baby!
I’m glad Max elucidated this phenomenon that I’ve been experiencing, but hadn’t quite put words to. I don’t really get stuck anymore. That’s agentic coding’s hidden productivity boost. And when I do get stuck, it doesn’t last long. Plus, I don’t spend that stuck time banging my head against a wall until there’s a hole large enough for me to slip a solution through.
I just let my agent tap into its vast repository of extant solutions and use my past experience to guide it toward the best choice. 🤌👨🍳💋