@paulnovosad Think a good chunk of SWEs also hate reviewing LLM code and prefer coding themselves, but feel like they're being pressured to use them?
Think it might be sort of inherent to the task a bit.
@paulnovosad ..actually work on them to ensure they develop the skills to score well on the summative assessment.
Idk if students hating something is a distinctive downside
eg https://t.co/UAMdFmL4Zn
@paulnovosad What are the sorts of situations where a student who's been doing all their assignments (of whatever form) would fail/do poorly even with AI access, while students who haven't been relying on AI so well?
And why can't those sorts of situations be turned into new assessments?
@paulnovosad ...students and have them report on the issues, potential improvements, review the code, identify any bugs etc.
This has a high fixed setup cost. But then marking should be somewhat straightforward if you can reference against your list of issues.
@Afinetheorem (implicit premise being that the ultimate political endgame from the status quo and populist backlash etc is actually going to be quite bad for the rich)
@Afinetheorem Alternatively is the issue that most of the cost of stuff like wealth taxes or populist measures actually falls on people who aren't rich, so the rich are actually underinvesting in addressing this?