There's an utterly ridiculous "study" out from Stanford about "ghost engineers" which are reportedly engineers who do nothing at companies.
-
@thisismissem do you have a DOI?
-
@DenialShown nah, it was a bunch of tweets by a person that worked on a pre-print that was assessing something else. The "pre-print" claim by this person appeared to mostly just be lending credibility to what he was saying, along with using Stanford's name for legitimacy
-
Hrefna (DHC)replied to Emelia 👸🏻 last edited by [email protected]
This matches my impression as well.
Is it too much to hope that his faculty adviser throws him out on his ear for this nonsense?
Oh wait his faculty adviser is a business and psych researcher who regularly goes outside of his lane.
Sigh.
-
@drahardja @blogdiva @thisismissem
Goodhart's Law strikes again!
https://en.wikipedia.org/wiki/Goodhart's_law -
@samir In Germany it would be invalid to fire someone w/o a warning. So the moment I'd get that warning judging my none performance based on not enough code changes I'd maliciously comply and do lots and lots of commits just to teach everybody the meaning of Goodhart's law by example.
@thisismissem -
@rockylhotka @thisismissem My favorite performance metric is actually LOC *removed*.
Quality over quantity, every time.
I'm lucky that I have a client who gets that and is also happy to allocate some budget to writing documentation.
-
@thisismissem there's probably a pretty interesting paper to write on how to remove bias and improve accuracy in selecting a group of humans to review someone else's code and impact..
But I can almost already imagine this one didn't waste time on that kind of scientific rigor
-
@hrefna @DenialShown will be interesting to see what the ethics committee says about this
-
@levampyre @samir not under Probezeit. Anyway, they fired the entire team I'd been assigned to a month later
-
@rune we came to the conclusion that the "10 person expert panel" likely wasn't, because the previous pre-print was based on whether LLMs could replace human code reviews.
-
@thisismissem Wait, are you/was the team employd in Germany? Because a mass reduction again has additional formal requirements that apply even if individual employees are on probation still. Anyways, if you or any of your colleagues needs advice on German workers rights and regulations, hit me up. I'm a works council member of 8 years and just dealt with that same shit in my own company. So my knowledge is as fresh as my thirst for blood.
@samir -
@levampyre @samir you're about 7 years too late for that, but I'm certain that the company would've "found" evidence to support their claim despite having me change my visa to them and go permanent from freelance but a month earlier.
-
@thisismissem Noooo! It's never too late to fight capitalist exploitation and abuse. Workers unite!
@samir -
@thisismissem What does "simulating a panel of 10 experts to evaluate each commit" even mean? Sounds like a concept of a plan, at best.
-
@thisismissem have they considered simply asking the engineers what they do?
-
@erikriffle why would you do that when you could write a garbage paper promoting LLMs? /s
-
@thisismissem @esk I would absolutely love to have more time to write code at work. I do so many other things though it’s hard to find projects where I can.
-
Kris Mitka :java: :rust:replied to Emelia 👸🏻 last edited by
@thisismissem they are overlooking the commits you talk others out of.
Something something … chalk mark…
-
Emelia 👸🏻replied to Emelia 👸🏻 last edited by [email protected]
Hey, so someone wrote an article about this terrible paper too: https://yieldcode.blog/post/ghost-engineers/
And @changelog issues a retraction in their latest newsletter about the veracity of the claims:
-
@thisismissem paging @jasongorman
