this post was submitted on 30 Oct 2026
852 points (99.2% liked)
Technology
89311 readers
3234 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 3 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
So, funny story that might be related. Engineers directed by their management to tokenmaxx often will just prompt very complex, difficult tasks to drive up their token usage.
Using the metric as the goal is a great way to give perverse incentives. SEE ALSO the US Stock Exchange.
Giving devs metrics and targets is just so dumb.
Their job is to learn the rules to manipulate systems to their benefit.
If you hand them some half-cooked system they're going to work out how to (best case scenario) bend it out of their way.
Worst case scenario is that they bend it to their benefit, and find a way to do less work for the same "score".
Everyone needs metrics and targets, otherwise it is just dicking around and nothing useful gets done.
The problem it that tech bros only half listened in social sciences class and choose the dumbest metrics and targets. Agile completely killed waterfall and specification development. In turn, using the equivalent of ignorance of the masses ticket systems on Jira. The worst of all worlds with commits, PRs and lines of code as metrics and a vacuous tick mark on user stories as goals. No wonder software development has been hell for decades.
Software developers need to stop designing in the code and go back to designing on paper against specifications and unit tests. Thinking has always been the main work of a software developer, not just writing code. As for the AI side of the problem, the purpose of code is not just to be written, but to be read and understood by humans. AI coding trashes the ontological purpose for the existence of code.
Middle managers like to set targets and goals because it counts towards their bonus, and it looks less like they've got their dick in their hand when the boss comes calling.
No one "needs" anything to do their job. A brief is helpful, a deadline is worth taking note of, but anything else is just interfering with people who are genuinely there to just do the job.
Most productive jobs i ever had? The ones where i was asked to do something and then allowed to do it.
As to the latter part of your message, it feels like you're just trying to process something so i think I'll just leave you to it. It's a bit too non-specific to respond to though.
Yeah, getting metrics on software development has always been a problem from the first time people were writing code in the employ of a business.
Businesses love metrics. Put them on spreadsheet makes some graphs, that's what some people do as their job.
The problem is you can't actually measure the productivity of someone that's doing something that's never been done before. Software development is always about doing things that haven't been done before. If you want the same software, you just copy it, you don't need a developer. You only need a developer to make the software do something it doesn't currently do, which means having someone do something they've never done before.
You just can't measure the productivity on that. But businesses want to have some numbers, so they chuck whatever numbers they have into a spreadsheet.
You might be interested in this (it was super timely for me 6 months ago) : https://leehanchung.github.io/blogs/2026/04/05/the-ai-great-leap-forward/
That science adage: To measure, is to change.
When a metric becomes a goal, it's no longer a good metric.