If I was a betting man I would say that security engineering is going to see a spike in demand in the next few years as a by-product of this.
But that’s probably cope.
And yet, the latest generation of models is surprisingly underwhelming.
Hard disagree. The latest generation (Astra 6/Opus 5.5) is the first one that doesn’t produce slop IMO. They basically fixed Claudish, the code quality (at least in Rust) is close to what I would write most of the time, they don’t spew a gazillion useless comments through the code, etc.
I don’t think you can talk about agentic coding without saying which models you’re talking about - there’s a huge difference between the latest frontier models and those from 6 months ago.
I feel the Team Fabric point so strongly. My team has gone from 2 reviewers per PR, to now 1, and open conversations about going to 0 with only AI review.
Feels less like a team every day, and more like a room of individuals, with less and less understanding and insight into what others are doing. Everyone is carving out their own kingdoms in the repo, that only they and their agent understand and can operate on, with internal patterns and designs.
Do we not see how big a risk this is if anyone leaves? Code review was for finding issues for sure, but for me that is at most half the purpose. The other purpose is driving consistency across the codebase, and disseminating knowledge amongst the team.
I feel that despite the stereotype of software engineers as loners, a lot of people I’ve worked with are deeply collaborative, and this is fuel on the burnout fire.
My honest, biggest, fear is that AI coding does not, in the long run, drive up bugs and security issues. That is the only tangible piece of data we have to point to as a counter argument to this tool. If not for that businesses will not care about consistency, only that the agents can read and understand it.
It’s crazy how little we as coworkers ask each other for help now. It used to be nearly daily for my team, could last 2 minutes or 2 hours but it was always an interesting moment
The sad thing is that AI coding does not have to be done sloppily, but somehow, corporate leadership has pushed us in this direction. They wanted some wins to report to wall street, and they had to sacrifice software quality.
The quality bar for software has been greatly reduced across the board, and normalized slop even outside of the corporate world and into open source software.
I wish for a world where AI coding capabilities still existed, but we did not tolerate slop as a society.
