How do we know this is due to AI usage? Perhaps it is because the students missed key in-person learning at the tail end of high school due to the pandemic lock-downs? I cant imagine learning calculus / linear algebra on my own in high school.
Failing grades soar with AI usage, dwindling math skills in Berkeley CS classes
461–470 of 828 posts
Re: Failing grades soar with AI usage, dwindling math skills in Berkeley CS classes
#462> “I’m a strong, strong opponent of what Harvard is doing to say that only a fraction of students can earn A’s,” Garcia said. “I think you should have clear standards for what an A means, and then give tons of opportunity for people … to get to that A bar without lowering the standard. So everybody who’s curving is hiding that effect. It’s completely hiding that effect, and it’s pretending as if nothing’s wrong, and…
I think curving has its place. One of my math professors explained that in his opinion an effective test should differentiate performance as much as possible. The top students should score very well and the bottom students should score very poorly. If all the scores are clustered near the top (>80% for example) then it's hard to tell who really mastered the material and who just muddled through. Then, once you've sor…
Yes, curving has a place here - and it is to evaluate, as you put it, whether the test differentiates performance as much as possible.
If you curve the students after the test, you are applying subjective edits to the graded performance just so the distribution of grades matches the measure of your tests effectiveness. That's just hacking the metric.
Further, even if you believe that tests should differentiate mastery (not students), your test should have teased out the differences or given you enough confidence to provide As to everyone who mastered the material - which should be absolutely possible! There's no a priori reason that all students cannot absolutely get the same grade, except for the a priori assumption that grades are for differentiation of students themselves (this year's A means this is the best student of this year), vs indicating mastery (all students absolutely crushed this exam).
You can dock points for style, or unnecessary struggle, or whatever subjective metric you want, but fudging the grades based on vibes to fit a prior-assumed distribution is just kinda "test effectiveness laundering"
Re: Failing grades soar with AI usage, dwindling math skills in Berkeley CS classes
#463I'm curious about LLM adoption by faculty. Is it possible that lesson plans and/or slides are being vibe-produced by professors/TAs, potentially reducing quality of instruction?
Re: Failing grades soar with AI usage, dwindling math skills in Berkeley CS classes
#464Earlier quoted context omitted.
Again you are not understanding the comparison. I’ve got 20 inch lean arms - I know far more about muscle building and retention than you. I train just as hard to maintain them as I did to get them there. The people who say “oh it’s easy to maintain” LOL it’s easy to maintain 16 inch arms.
> I know far more about muscle building and retention than you I am a competitive bodybuilder… > I train just as hard to maintain them as I did to get them there. Are you enhanced? Were you enhanced when you built the 20” arms? If so, yes I agree. Edit: With 20" arms, there's nearly 0% chance you're natural. You can't compare your enhanced experience to naturals.
Chiming into this little tiff to say I think bulk muscle is a bad analogy in the first place. It’s more akin to a muscle memory/skill. Something like golf is a better analogy. If you took any golfer, at any level, and had them refrain from golfing for 3 years. I feel pretty confident asserting they would all perform worse than they had. Their skill is diminished.
Re: Failing grades soar with AI usage, dwindling math skills in Berkeley CS classes
#465Earlier quoted context omitted.
>We used a vibes based system based on how compelling a sob story you could concoct whilst staying on the good side of fraud. Besides lost meritocracy, that is accidentally filtering for ability and willingness to manipulate others emotionally. Which feels really scary.
Honestly I don’t believe a thing they’re saying. Just look at the demographics of any Ivy League school and you’ll see it can’t be true
My kid applied to Brown and others and the prep company we hired ($150/hr) spent more time going through the "sob story" the parent talked about than any other requirement (kids scores are to be fair quite decent already).
Re: Failing grades soar with AI usage, dwindling math skills in Berkeley CS classes
#466This is concerning. AI tools are powerful but foundational skills still matter. The balance is tricky for CS students.
Re: Failing grades soar with AI usage, dwindling math skills in Berkeley CS classes
#467Earlier quoted context omitted.
No matter how many times I read it, I can't interpret it the way you're suggesting. "x soars after y" always reads as "x increases a lot because of y". I don't really get what you're saying. Are you maybe saying that "soars" might mean "get better", so "failing grades soar" might mean there are actually less failing grades? That's not how I've ever understood that word.
"Falling" means that something goes towards the earth. "Soaring" means the opposite. "Grades soar" means that grades went up "Falling grades means that grades are going down". "Falling grades soar" is just meaningless writing.
The title is "failing grades soar" (one 'l', not two), not "falling grades soar."
Re: Failing grades soar with AI usage, dwindling math skills in Berkeley CS classes
#468Earlier quoted context omitted.
Again you are not understanding the comparison. I’ve got 20 inch lean arms - I know far more about muscle building and retention than you. I train just as hard to maintain them as I did to get them there. The people who say “oh it’s easy to maintain” LOL it’s easy to maintain 16 inch arms.
> I know far more about muscle building and retention than you I am a competitive bodybuilder… > I train just as hard to maintain them as I did to get them there. Are you enhanced? Were you enhanced when you built the 20” arms? If so, yes I agree. Edit: With 20" arms, there's nearly 0% chance you're natural. You can't compare your enhanced experience to naturals.
[flagged]
Re: Failing grades soar with AI usage, dwindling math skills in Berkeley CS classes
#469Earlier quoted context omitted.
I mean the argument that is being put forward is that it isn't a pneumatic nail gun for the human mind - it atrophies are mathematical capability and quality of understanding.
Right - but you could imagine a similar sentiment anytime a skill was replaced by a new technology. I think the jury is still out on whether LLMs actually lead to complete atrophy of skills that don't eventually get replaced with brand new skills.
Anecdotally people have been noticing atrophy quite a bit. Again its anecdotal because we can't possibly have a study that works in real time because the technology is rolling out insanely quickly.
And all the older technologies that have rolled out haven't competed against our cognitive abilities at speed and scale.
I don't think of cognitive ability as a skill per se - more of a critical core function of humanity.
I say this as someone who uses it extensively not some luddite but is also very aware of the risks which I assume are worse for people who have limited understanding on the matter.
Re: Failing grades soar with AI usage, dwindling math skills in Berkeley CS classes
#470I have some sympathy for these kids. If LLMs were around when I was a student, I would've also used them to "speed up" my homework assignments then proceed to fail all my tests. Now I work mostly with PhDs who were at the top of every academic environment they've ever been in. And yet I can see their thinking skills rapidly declining as well; many of them can no longer brainstorm, code, think deeply, or write without…
Counterpoint, I think this is true for some archetypes of people, but certainly not everyone. I personally use it like the socratic method. I am an intermediate user, I spend a ton of time with LLMs at work and personally, both prompting and letting some crappy agents try to automate boring work. I primarily use Gemini and ChatGPT models, along with some Chinese smaller weight models (eg qwen) locally. If you treat t…
This, I will use Obra Superpowers brainstorming skill to propose/refine a few viable solutions for a feature or bug I'm trying to solve. After it asks me clarifying questions and presents a spec, I will say "well what about X or Y". The I'll run the grill me skill on the spec to tighten it up, clarifying any assumptions made.
I find it to be a really tight loop and results in very high quality code at a high velocity.