Post

My AI enthusiasm ebbs and flows. Today I’m realizing that some of my perceived productivity increase is an illusion. I merely moved some of the effort from writing code to having to micromanage a coder.

4 reposts14 likes

Writing the code is the fun part. Micromanaging interns who can't learn or grow is horrible. I genuinely can't understand why anyone would make that trade.

I guess a lot of the boosters would say that solving problems was the fun part, and you're just solving problems at a higher level. Like writing C# instead of ASM, now you're writing prompts instead of C#

I may not find literally writing code as fun as others do, hard to say, but what was genuinely fun for me and lost was helping other people grow. While I still had human teammates, this was falling off a cliff as everyone unblocked themselves with AI instead of reaching for help from other humans.

I was never the sort of dev who was a prolific producer of code. I have even mused that my biggest contributions were preventing bad code from being written and not writing good code.

Curious: How do you start a feature? Do you just describe the feature and tell it to go? Do you have it interview you? Do you write down a spec or plan? Something else?

We usually have a long discussion, plan mode, then execute. But the discussion can feel like micromanagement too because Infind I have to spell a lot more out to it to get a good plan than I would get from an experienced colleague.

I decide generally, but I’ve noticed the agent do a welcome thing and spot something we should do next. I’ll say “remind me about this when we’re done with this” and then when the reminder comes to close the session, I’ll say “draft that as a prompt for the next agent.

And an inexperienced colleague is becoming an experienced colleague when they did the same with me. With the agent, I have to keep diagnosing why the lessons we wrote down aren’t helping them do better the next time. It’s like debugging markdown and that’s harder than debugging code!

Yeah this is something I struggle with a lot. Giving feedback to a junior engineer "sticks" while it doesn't with agents, even with memory and updated instructions and stuff it's always a crapshoot whether it'll just do all the same stuff you had to correct last time

I’m now at the point where I will attempt Fable for everything which will blow through my $100 Max plan but if it improves or eliminates the micromanaging I need with Opus, may well be worth $200.

(And yes I know these Max plans for individuals are insanely subsidized, but it’s still an open question which is best value for me as they stand.)