Easy button problem
I am lazy by default and I am aware of it. Walk through the house, see a toy on the floor and the default is to ignore it. I have to tell myself “don’t be lazy” and pick it up. It is very easy to ignore it and revert to the default. Like an addict.
Heavy AI agent adoption is happening in my workplace. I mostly enjoy the benefits. Make a task narrow enough, point it in the right direction and get a pretty much done product. The job has moved to identification and planning + reviews. A lot of them, unfortunately. Review what the harness is trying to spin, verify outputs, check other team members a bit more carefully than usual as you are aware they could fall into the same lazy comfy routine. I think we did not become this to be code reviewers. We like hard problems, picking them apart and slowly cracking them piece by piece with the “power of coding” until they are solved.
That same agentic AI adoption prompted several of my panic moments where I was not sure if I even still knew how to code. Had to solve some tasks on Advent of Code to convince myself. The most recent moment prompted me to create a toy programming language.
It was very English-based and the target was JS since I’m most familiar with it.
create function named hello that prints mario.
call hello.Then build a compiler, learn what a tokenizer is, an AST, rebuilding the AST to JS. Made me appreciate some of the languages much more. My compiler actually spits out:
const hello = () => console.log('mario');
hello();But what worries me more is the number of moments I thought: “I can just give this part to Claude.” Easy mode is right there to embrace you into sweet laziness.
Lazy as a SaaS.
I have been trying to structure my thoughts on this whole new world for some time now and failing to do so. Who am I in all of this, what’s the point, is this really a job now and in the future. Maybe I am too lazy again to consolidate my thoughts, but this time I am not even aware of it.