Failing CS Grades Soar At UC Berkeley As Professors See Greater AI Usage (dailycal.org) 110
The University of California at Berkeley discovered the percentage of failing grades in multiple CS classes this spring "is significantly higher than past semesters," reports the campus's student newspaper.
"Instructors point to students' increased reliance on AI, lack of mathematical preparedness and understaffing as potential contributing factors." According to [coursework platform] Berkeleytime, 35.3% of CS 10 students and 10.6% of CS 61A students received F's in spring 2026. In spring 2025 and spring 2024, the percentage of F's did not exceed 10% for either class. The electrical engineering and computer sciences department's grading guidelines state that 7% of students in lower division courses, including CS 10 and CS 61A, should receive D's and F's...
[UC Berkeley teaching professor Dan Garcia, who taught both classes] believes the "primary driver" of these abnormally high failing rates is due to a "vast increase in academic dishonesty" due to students' usage of large language models, such as Claude, ChatGPT and Google Gemini. "Some of the numbers that you saw from the number of students who receive failing grades were because we caught them (cheating) and prosecuted them and are sending their cases to the Center for Student Conduct," Garcia said. "But in other cases, it's students who are leaning a little too hard on LLMs to do their work for them, and then at exam time just really aren't ready." According to Garcia, nearly 30 students in CS 10 were "caught cheating on take-home exams" in spring 2026...
In addition to overreliance on AI, Garcia also pointed out that many students are underprepared mathematically, a concern echoed by campus associate teaching professor Gireeja Ranade. Ranade noticed a similar lack of prerequisite mathematical skills in her spring 2026 EECS 127 class, "Optimization Models in Engineering," which she described as "differently challenging" to teach this semester. The class saw a 16.8% F rate, far higher than the 5% of D's and F's that the EECS department describes as "typical" for an upper division course...
Both Garcia and Ranade have joined more than 1,300 UC faculty in signing a petition calling for the reinstatement of ACT and SAT standardized testing scores for STEM admissions in the UC system.
Thanks to long-time Slashdot reader theodp for sharing the article.
"Instructors point to students' increased reliance on AI, lack of mathematical preparedness and understaffing as potential contributing factors." According to [coursework platform] Berkeleytime, 35.3% of CS 10 students and 10.6% of CS 61A students received F's in spring 2026. In spring 2025 and spring 2024, the percentage of F's did not exceed 10% for either class. The electrical engineering and computer sciences department's grading guidelines state that 7% of students in lower division courses, including CS 10 and CS 61A, should receive D's and F's...
[UC Berkeley teaching professor Dan Garcia, who taught both classes] believes the "primary driver" of these abnormally high failing rates is due to a "vast increase in academic dishonesty" due to students' usage of large language models, such as Claude, ChatGPT and Google Gemini. "Some of the numbers that you saw from the number of students who receive failing grades were because we caught them (cheating) and prosecuted them and are sending their cases to the Center for Student Conduct," Garcia said. "But in other cases, it's students who are leaning a little too hard on LLMs to do their work for them, and then at exam time just really aren't ready." According to Garcia, nearly 30 students in CS 10 were "caught cheating on take-home exams" in spring 2026...
In addition to overreliance on AI, Garcia also pointed out that many students are underprepared mathematically, a concern echoed by campus associate teaching professor Gireeja Ranade. Ranade noticed a similar lack of prerequisite mathematical skills in her spring 2026 EECS 127 class, "Optimization Models in Engineering," which she described as "differently challenging" to teach this semester. The class saw a 16.8% F rate, far higher than the 5% of D's and F's that the EECS department describes as "typical" for an upper division course...
Both Garcia and Ranade have joined more than 1,300 UC faculty in signing a petition calling for the reinstatement of ACT and SAT standardized testing scores for STEM admissions in the UC system.
Thanks to long-time Slashdot reader theodp for sharing the article.
So let them fail (Score:2)
Understanding what an LLM can and cannot do should be part of the knowledge of a CS student, they clearly don't so deserve the failing grade.
Re: (Score:3)
Dear Claude,
Please outline an algorithm that you currently cannot solve that human intuition could implement within a class assignment.
Optionally write a flawed test harness for the algorithm that if implemented as proposed would catastrophically fail the task. ...
Re: (Score:2)
"I am literally still using the black and white brother laser printer I purchased on my kid was in grade school and they graduated college almost 5 years ago so my kid went for a STEM degree"
Count your blessings; you've got a great printer, that won't be trying to rip you off with over-priced OEM toner / ink, monthly subscriptions, built in obsolescence. Trust me, you do not want a new one.
Re: So let them fail (Score:2)
ChatGPT says there are students that learn and students that coast. AI is just another tool that spans the difference.
Re: (Score:2, Insightful)
ChatGPT says there are students that learn and students that coast. AI is just another tool that spans the difference.
AI is a tool that increasing the coasting.
Re: (Score:3)
Throwing those that coast out is a good thing. Bad engineers are a cost to society, not a benefit.
Re: (Score:1)
Your comment presupposes that those who coast will just get washed out and it will only affect them. Instead:
(1) As a result of AI, students who would have invested the effort and become solid developers will instead "coast" thru.
(2) Many of the coasters will still get their degree, enter the job market, and obtain developer positions.
They may even have an advantage over non-coasters because they will have had more time to devote to extracurriculars and internships.
NET RESULT: AI weakens the overall quality
Re:So let them fail (Score:5, Insightful)
Understanding what an LLM can and cannot do should be part of the knowledge of a CS student, they clearly don't so deserve the failing grade.
A student should understand what an LLM can and cannot do, but a CEO firing humans by the hundreds to replace them with premature ToddlerAI, somehow gets a pass?
Make it make sense. Because Greed is destroying the point of a student. And a classroom.
Re: (Score:2)
Because Greed is destroying the point of a student. And a classroom.
That has a tendency to self-regulate longer-term. We are currently seeing one really big fat beautiful example of that.
Re: (Score:3)
Because Greed is destroying the point of a student. And a classroom.
That has a tendency to self-regulate longer-term. We are currently seeing one really big fat beautiful example of that.
In every other human revolution across all of known history, the answer for the just-made-unemployable buggy whip maker was go get an education. Go re-train yourself. Your job no longer exists.
This revolution, looks to replace the human mind.
Which ultimately means any tendency humans might have to suggest tendancies, patterns, leanings, or even proven historical changes become null, void, and bullshit. Because all it now takes is Good E. Nuff AI to essentially destroy human employment. Along with peac
Re: (Score:3)
This revolution, looks to replace the human mind.
Yes. Or at least it gets marketed as such and people are generally too dumb to understand that this claim really is 95% hot air.
In any case, this makes things fundamentally different. And to reach that 20% unemployment (which may indeed be enough), this tech does not even need to deliver. It just needs to be around long enough and too many utterly dumb C-levels need to believe the lies. Ido not think we will see that this time though, the business numbers are too abysmally bad and we see more and more of th
Re: (Score:2)
If people "generally too dumb" end up being CEOs, VC fund managers, and lawmakers, then it doesn't really matter if the claim is hot air or not. Even if AI is 95% hot air, the replacement theory of firing tens of thousands justified by an undeniable Recession doesn't exactly paint a rosy picture on the told-you-so alternative storyline.
Hard to really sell the dismissal of this problem with those of us who clearly remember the social and financial impact of dot-bomb vaporware. Sure would be nice if we coul
Re: (Score:2)
Well, we will see. But it really is only timing that could still save things to some degree. And timing is hard to predict.
Re: (Score:2)
These students see the most successful people in the world advocating for this. Throwing away any real skill in favor of buying tokens. Much of the social/media is calling the same tune.
New students are being steeped in this before they ever set foot inside a college. That gives the educators an uphill battle. It's going to give employers a bit of a situation, too.
Re: (Score:2)
Will anyone run to the rescue of these C*Os, when they fail? Many of them will, and nobody will shed a tear for them. Contrary to this the students probably expect someone to solve their now upcoming difficulties for them.
Re:So let them [learn to collaborate with AI] (Score:2)
Hmm... Only "insightful" comment in the FP branch? Maybe the moderators have been replaced with genAIs?
(And no Funny anywhere, as expected.)
As regards the story I'm remembering a recent MIT video. Long section about how to make AI work with the course. On the negative side, it recommended an anthropomorphic approach, basically treating the AI as though it were a human collaborator, but with "usage limits" to keep it in a subordinate status. On the positive side, I forgot. Maybe I should ask an AI for help?
I
Re: (Score:2)
Make it make sense. Because Greed is destroying the point of a student. And a classroom.
And life.
Re: (Score:2)
A student should understand what an LLM can and cannot do, but a CEO firing humans by the hundreds to replace them with premature ToddlerAI, somehow gets a pass?
We all know CEOs firing people in favour of implementing a more expensive and shittier option are idiots - I don't know who's giving them a pass other than other CEOs doing the same thing. It's a giant circle jerk of jerks.
cull the weak (Score:3, Insightful)
You mean rip-off the weak. (Score:1)
We're a litigious society, I wouldn't blame those that failed to demand their money back due to inadequate entry rejection testing.
cull the weak (Score:2)
A five-fold increase in course failures, mostly attributed to cheating, would not be lessened by requiring a minimum ACT or SAT score before entering the program.
Filing a lawsuit that says "The school didn't check if I was smart enough to pass the two-digit entry-level classes so I felt justified in cheating" probably won't get far.
Re: (Score:3)
Re: (Score:3, Interesting)
It's not meant to help students. It's meant to get more butts on seats paying tuition to keep the staff in cushy, well-paid jobs.
No-one cares whether the students pass or fail except to the extent that it affects next year's revenue. And kids who would legitimately qualify under the SAT scores are smart enough to realize that taking a degree is a very bad financial decision these days.
Re: (Score:2)
The removal of those tests helps those who have not been tutored to pass those specific tests ANS who also have the natural aptitude that will allow them to excel when they enter higher education. I've met students like that. The overall number may be small, but the number of students who benefit from the removal of that hurdle is non-zero. It's up to you to decide whether to treat them as acceptable losses.
What we actually need is a better method of selectively assessing who is ready for higher education.
Re: (Score:3)
A five-fold increase in course failures, mostly attributed to cheating, would not be lessened by requiring a minimum ACT or SAT score before entering the program.
Filing a lawsuit that says "The school didn't check if I was smart enough to pass the two-digit entry-level classes so I felt justified in cheating" probably won't get far.
If the cheater actually got vetted and denied in a valid ACT or SAT test that required a minimum score unaided by AI, your excuse to cheat in a college course would have never manifested in the first place.
Re-institute mandatory ACT/SAT scores as a proven vetting measure for college. Just because they can afford to treat students like idiot customers doesn't mean they fucking should.
Re: (Score:2)
If you aren't required to pass say an SAT or ACT test for entry then you are doomed to waste your own money by failing - which isn't your fault if some greedy school disabled the test to let you in so you can fail while they get paid. We're a litigious society, I wouldn't blame those that failed to demand their money back due to inadequate entry rejection testing.
That same litigious society already sued on the basis of mandatory testin, er I mean racist practices designed to deny higher education to intellectual minorities.
Why do you think the tests got disabled.
(Yeah. I know. Hell of a Catch-22 to swallow.)
Re: cull the weak (Score:4, Interesting)
Re: cull the weak (Score:4, Interesting)
Academic educator here. I do not think this is accurate. The weak ones sometimes have problems that can be fixed, like poor impulse control (and that must be fixed or they will not be qualified at the end of their studies), but mostly they want the grade for cheap and would go on being bad engineers and hence a massive cost to society.
STEM subjects either need to limit enrolment ("soft", like Mathematics by reputation, or "hard" by aptitude tests) or cull mercilessly. These subjects are very much a case of "most people cannot do it well or at all" and letting them graduate nonetheless is a very bad idea.
Re: cull the weak (Score:4, Interesting)
This is the biggest "generational divider" I see rapidly creeping up over the last 5-10 years: People want "The Bag" but don't want to put in the effort to get it. They think the journey and the effort to get there is just in the way, and once they get their Bag, then they'll put in the effort. But it doesn't work that way. You can't flip on Effort like a switch if you've been half-assing things for a decade leading up to that.
Obviously, there are a lot of factors at play as to why the percentage acting that way are going up: K-12 schooling being mostly concerned with keeping grades up to retain funding so they pass everyone, abolishment of ACT/SAT scores for college entry as mentioned in this story, being distracted by the over-use of mobile devices/laptops in class, overuse of AI now so they once again don't really understand the material, etc. Not to leave out: Feeling like society is rugpulling them, so why bother trying hard when you're going to end up unemployed or underemployed no matter what you do?
As for why they lean into AI: I don't think we should discount the general lack of tech literacy these days as a reason they fall into the AI landmine. Kids don't use technology so much as consume it. The knowledge has never been easier to find and learn how to understand tech, but most simply... don't. It's not uncommon to have students entering college that don't understand what a filesystem is and how to navigate it for saving files, etc. In which case it's no surprise they fall for AI when the AI companies go out of their way to call LLMs "intelligent".
Re: (Score:3)
This is the biggest "generational divider" I see rapidly creeping up over the last 5-10 years: People want "The Bag" but don't want to put in the effort to get it. They think the journey and the effort to get there is just in the way, and once they get their Bag, then they'll put in the effort. But it doesn't work that way. You can't flip on Effort like a switch if you've been half-assing things for a decade leading up to that.
Well, let's take a closer look as to exactly how Gen LazyBag got that mentality, shall we?
The most popular way to get-rich-quick in the 21st Century, is to become a professional social media narcissist who is good enough at getting an audience to watch monetizing content on the regular. Unfortunately, only about 0.0001% of YouTubers actually get what the 21st Century would define as The Bag (seven figures or more, according to hypergamy standards), but let's take a closer look at not just a rich YouTuber b
Re: (Score:2)
This is the biggest "generational divider" I see rapidly creeping up over the last 5-10 years: People want "The Bag" but don't want to put in the effort to get it.
I am not seeing it in my students or not more than in the last 30 years. But these are all advanced (and the weak ones have already washed out), I am teaching in Europe and I am teaching subjects that are known to be more difficult (IT security) and that are elective. So my view may be skewed.
Re: (Score:2)
In the end, it was found that the original marshmallow test didn't deconvolve impulse control from trust in authority. If the students don't trust that AI won't be used against them, e.g. faulty AI checkers used to accuse them of AI use, their non-AI skills not valued on the job, etc., then it is rational for the students to use AI.
Re:cull the weak (Score:4, Interesting)
I have seen numerous headlines saying that the corporate world is blaming AI for layoffs as it gives them an easy scapegoat. It sounds like this is something similar.
It would be one thing if they said students were failing courses due to not understanding key concepts of computer science supposedly learned in class and demonstrated in projects or labs. That might easily be explained on overreliance on AI tools.
But not having the required math fundamentals required for a course? That's not on AI that's on them for not vetting the incoming students. And then to throw in the reminder that the UC system stopped using test scores for admittance, well that seems to be the nail in the coffin.
I actually went to a UC school a little over 20 years ago (not Berkeley) and I remember my high school friend and I both had the plan to enter the CS department together. My friend had much better grades than I did, including math. However, based on our SAT/ACT scores he had to retake calculus and take an extra English class that I didn't, and it ended up being prophetic as he struggled with CS (particularly the math) and ended up switching majors.
Re: (Score:3)
But not having the required math fundamentals required for a course? That's not on AI that's on them for not vetting the incoming students. And then to throw in the reminder that the UC system stopped using test scores for admittance, well that seems to be the nail in the coffin.
They found out denying Stupid entrance, wasn't nearly as profitable.
Profits also started to soar when they normalized a college campus as a valid lifestyle choice for those choosing to hide from Reality for the better part of a decade. That's not a failing student. That's a high-paying customer who merely needs a schedule adjustment.
It's ironic the ones acting like grown-ass children also demand to have an Right to Vote. Hell of a way to convince society we should raise the voting age. Can't imagine alc
Cull the Profit Motive. (Score:2)
That is one of the two jobs of college - to winnow out the retards so the degree means something for the rest, and to educate those who can learn.
Actual valid entrance exams failing them before day one would be a valid mitigation. But then colleges wouldn't be able to string along their high-paying customers for tens of thousands of dollars in profit first, only to be told later that they're not really cut out for college.
And by "later" I mean never. Since the only thing colleges seem to care about is the check clearing. You can now be a perpetual student for decades if you really want to. Some never leave the campus bubble. We call them "admin
Re: (Score:2)
can't cheat your way to knowledge (Score:3, Insightful)
This isn't The Matrix and you can't just download new skills. You have to do the work. Who would have guessed?
Re: (Score:2)
This isn't The Matrix and you can't just download new skills. You have to do the work. Who would have guessed?
30 years from now your personal WorkerBot will be downloading the skills that replaced your weak-ass at-human-speed job a decade prior.
There will be two kinds of workers in the future. Those who can afford a premium WorkerBot to do their mental and physical work for them, and those who will BE the worker in a leftover profession soon to be retiring meatsacks permanently.
Our modern world, made Orwell a damn prophet. The one thing we've proven to be bad enough to be good at, is grossly underestimating a fut
Re: can't cheat your way to knowledge (Score:2)
Re: (Score:2)
You really think that you're going to able to outsource your job to a robot while still staying employed? The more likely scenario will be that you will be replaced by the robot and you'll find yourself unemployed.
Lack of math skills? (Score:5, Interesting)
Looking that course up it seems that's about proving algorithms and other stuff, so maybe you do need more advanced math skills for such a theoretical course. My 39 years of experience is that 2 years of high school algebra is good enough for the typical programmer. I've worked 2 jobs in my time where higher math was needed, but the employer hired someone with a Phd in math to figure out what needed to be done, the rest of us did the UIs, the DBs, the infrastructure, etc. I'm not dismissing higher math, there are places where it's useful, but most programmers don't need it. Hmm, I also see that course isn't required to graduate, so some of those students shouldn't be taking it if they're not sure of their skills/knowledge.
Re: (Score:2)
Re: (Score:3)
IMHO "Calc 1" is worthwhile to learn if you're in CS. You get an introduction to differentials and integral Calculus, and there is some everyday-relevant physics equations to be learned related to position/speed/acceleration, etc. Anything beyond that is only going to have relevance in more niche jobs. In which case, either take the extra semester or two of Calc, or book up on it on-the-job since you should understand the basics already.
Re:Lack of math skills? (Score:5, Insightful)
Students who are unable or unwilling to learn these things aren't going to be able to learn CS because they lack the foundation(s) required, and thus they're likely to receive low grades. That's how it is, and that's how it should be.
This is not to say that people who only want to learn to program should not do so: they most certainly can. But that's a very different educational path than trying to learn CS. It's roughly the same as someone who wants to learn to be an electrician vs. someone who wants to earn a degree in EE.
Re: (Score:2)
Unfortunately employers tend to prefer people with a MsC rather than people that made vocational schools, and then find that people just graduated don't know the latest technology they use.
2-year community college rather than 4-year (Score:1)
Re: (Score:2)
Computer Science is about the science of computing - algorithms and other things. It's the same as studying physics, chemistry, and other similar sciences - providing the future theoretical foundations. Here a Turing machine can exist properly.
Computer Engineering is the about application of science - and like other engineering like civil, mechanical, electrical, etc. You have to deal with real-world compromises - just like not all cows are perfect spheres, you have to consider actual aerodynamics of your b
Re: (Score:2)
Re: (Score:2)
How much of "The Art of Computer Programming" by Donald Knuth is programming and how much looks a lot like pure math? What percentage of Computer Science grads can even read and understand book one?
We need two degrees. A vocation "programming" degree and college/university degree in "Computer Science."
Re: (Score:2)
It's roughly the same as someone who wants to learn to be an electrician vs. someone who wants to earn a degree in EE.
Electricians can be trained. EEs either know their stuff or get sent packing. Too arrogant to be (re-)trained.
But now the government needs to eliminate these lowest-bidder-sacks-of-crap EEs who rubber stamp Facility Blueprints created by overseas Techs (illegal, yet still done).
Re: (Score:3)
Re: (Score:2)
The problem is that there doesn't exist a degree that meets the characteristics of someone who wants to be a programmer (or non-research-related technical practitioner). I always hate when people trot out that the ivory tower of CS was meant for big thinkers and problem solvers, not programm
Re: (Score:2)
I worked (embedded systems) with no degree for 30+ years then decided to complete a degree. During my no-degree era, I did one integral that was offered by a manager as a voluntary exercise to anyone who wanted to try it. Years later, my project required a DPLL with digital filter. A digital filter specialist did the math, and I did the implementation, then modified it to put an acquisition/tracking gear shift in the filter to get rapid frequency lock, followed by tight phase lock. In a later project, I ca
Re: (Score:1)
Says you.
Computer Science is still science. When I finished my CS degree I only had to take 1 extra class to get a math degree.
If you work with graphics, games, CAD, AI, financial, etc, etc, all rely on math. If you're just a web "programmer" they yeah, maybe you don't need math. In fact you don't need a degree at all if you're just a web developer or business person.
Re: Lack of math skills? (Score:3)
Computer science is math. You don't need a computer science degree if your vocation is software developer. When I lived in Switzerland, software developer was a 2 year technical school degree. Computer science was a 4 year university degree.
Re: (Score:2)
1 year Software Engineering MSc for me, to turn the coding skills I'd learned as a physicist into something more formal. Project management and design skills, and then a broad spectrum of taster courses in languages, technologies, and topics - just enough to get you started if you went in that direction.
Re: Lack of math skills? (Score:2)
My CS degree had a lot of theorem proofs in it, invariants, that kind of thing. I've always had the habit of aiming to prove my code correct under all possible circumstances. Usually not a formal proof, but using the same skeleton as a formal proof would.
It got me a job on the C# language design team (when I tried to prove an algorithm correct, couldn't, discovered a counter-proof that the runtime had a flaw).
As I mentor junior devs and review their code, I'm always telling them to reason about their invari
Re: (Score:2)
Now in the age of AI, I find that invariant-heavy and proof-heavy guidance to the AI ends up getting its work done quicker and higher quality.
How do you do that?
Re: (Score:2)
[invariant-heavy and proof-heavy guidance to the AI] How do you do that?
My main AGENTS.md has ten lines about the most important coding principles:
- Prefer functional-style code, where variables are immutable "const", there's almost no "if/else" branching branching, and most functions are side-effect free. ...
- Code should have comments, and functions should have docstrings. The best comments are ones that introduce invariants, or prove that invariants are being upheld, or indicate which invariants the code relies upon.
I am adamant about clean engineering. What I look for: ...
- Invariants are the best way to document all aspects of code. These include code invariants (stating what assumptions a function makes about shared data, and how it upholds them), and architecture invariants (for instance the main index.js never touches state except through component accessors).
You must document *meaning* of every field, and also enums and disjoint type fields. ...
- "Meaning" says briefly what the field/enum represents. From a well-written meaning, a smart reader will be able to deduce all the invariants around this field/enum, and deduce how it will be used in the code.
- It is hard work to distill a good meaning! You must put considerable effort into it.
The instruction on "meaning" ended up carrying a lot of weight to the AI. It adopted the habit of putting a comment on every single field and function that starts with the word "// Meaning: " and they're honestly, genuinely good ones! Single-line sentences on fields that carry a lot of good weight.
Separately, I have a LEARNINGS.md file which I have the AI auto-update every time it gets course corrected by me. Over the first two weeks
Re: (Score:2)
Re: (Score:3)
Of course, there are other ways to develop that mindset as a programmer without going to college, but taking some math classes where you prove things is a really convenient/efficient way to do it.
Re: (Score:2)
There is no doubt that the mathematical mindset makes people better programmers. It trains you to think of all the edge cases, for example.
Of course, there are other ways to develop that mindset as a programmer without going to college, but taking some math classes where you prove things is a really convenient/efficient way to do it.
Your first sentence is correct, on average.
I bolded the parts which I do not think are proven. I believe that the real explanation for those portions is a combination of sample bias and reversed correlation-causation error.
I think the only claims we could confidently make about this issue are:
-The mathematical mindset makes it easier to program (ie write procedural instructions which rely heavily on basic algebraic expressions, ie math).
-People who have a mathematical mindset are more likely to enter math/s
Nah, Brah. (Score:2)
I'm not exposed to any programming that really needs to perform any extensive mathematics. But, I've seen a lot of code from those with and without higher math knowledge, degreed or not.
The code from those with higher math knowledge is almost always tighter, more elegant/efficient. Just clearly superior. The number of times I see counters and nested if-chains where a single modulus test should go is ridiculous.
It's not that they don't know about the modulus operator or how to use it. It's simply that they d
Re: Nah, Brah. (Score:2)
Standards, not gaussians (Score:4, Insightful)
It's perfectly fine to have standards. And those who don't meet them should fail.
But on the other hand, those who do meet them should pass. So, I was surprised to read this in TFS:
The [Berkeley] electrical engineering and computer sciences department's grading guidelines state that 7% of students in lower division courses, including CS 10 and CS 61A, should receive D's and F's...
Not if they meet the standards of the course. Look, I know it's unlikely that you'll get an exceptional lower-division class where everyone is deserving of a pass. But it can happen, and "grading guidelines" should not force failing grades on people who met the course requirements.
Re: (Score:3)
Re: (Score:1)
You're confusing departmental grading guidelines with some "course standards" that you've invented. Berkeley is saying that these courses should be difficult enough that 7% will get a D or F, and that's entirely reasonable. It's really the university's way of putting the onus on professors to make the courses sufficiently challenging, which they should be.
When I was a student at Berkeley the failure rate in courses was typically higher than 7%. I remember taking a physics course where about 30% got an F. I
Re:Standards, not gaussians (Score:5, Interesting)
My point stands. And I'm not confusing anything. The material covered in a course, and the level at which it is covered, are set before students enroll in the course. The pass/fail rate is determined by the kinds of students who enroll. It should not be set by departmental grading guidelines. How is a department to know whether a specific batch of students in a class in any quarter or semester will find the material difficult or not? They can base expectations for a specific course on historical statistics, but it's misguided to stipulate the percentage of fails by fiat as a supposed way of controlling how "difficult" a course is.
A passing grade in a course should indicate that you obtained competency with the material presented, at the level defined in the course description. You can't stop students from taking a difficult course they're not prepared to handle. Nor can you stop students from taking easier courses that they will ace because they're over-prepared. Pre-med students do this all the time, for example.
As for the topic in this story: failure rates are up because students are abusing AI tools and are ill-prepared for the course material. Sounds to me like the problem is a lack of adequate preparation for the incoming students. Prerequisites and/or skill-testing should be required of students before they can enroll in these courses. Perhaps the department would rather adjust the content of a course in order to keep a steady pass/fail rate. I think that's the wrong approach, because it just moves the goalposts, rather than giving students credentials for scoring goals on a standard playing field.
Re: (Score:2)
I don't think he misunderstood your question. The stats are not about judgement of the students. They're about the teaching ability and the connectivity of the material to be taught. Designing the courses, deciding on what materials goes into each course. There is always change.
If successive courses don't align well then students are going to be dropping out due to the knowledge gaps they aren't even being taught. So, using stats to monitor effectiveness is valid.
Re: (Score:3)
I don't think he misunderstood your question. The stats are not about judgement of the students. They're about the teaching ability and the connectivity of the material to be taught. Designing the courses, deciding on what materials goes into each course. There is always change.
If successive courses don't align well then students are going to be dropping out due to the knowledge gaps they aren't even being taught. So, using stats to monitor effectiveness is valid.
20 years ago they started a “standard” in American education; No Child Left Behind.
Today, illiterates ironically armed with a college degree are being fired at an alarming rate. Because employers need valid graduates. Not grown-ass Children who were never Left Behind when they absolutely should have been.
Maybe we should stop trying to assume the education should be dumbed down to “align” with the students. Because we end up treating them more like high-valued customers instead.
Re: (Score:2)
Re: (Score:2)
The problem is not the goal. No Child Left Behind is perfectly reasonable. The issue is that there two different approaches to achieve this goal: you can lower the standards (cheap) or you can give stragglers the means to success (expensive). Guess what the typical politician will do.
Lowering the classroom standard to the lowest common denominator, does not create success. It creates a standard of mediocrity that races to the bottom as a goal. When society's perpetual answer to idiot-proof is to build a better idiot, it's not perfectly reasonable to placate that.
And if we think throwing money at a moron is a "means to success", don't dare ask lottery winner bankruptcy statistics to validate that.
When it becomes undeniable that the current generation of students is less intelligent and
Re: (Score:2)
Re: (Score:3)
The outcome of education is to a very large part depend on the financial means of the parents. All too often that "moron" is the child of overworked parents without money to pay for tutors. Have you even considered that as usual many factors affect the school systems and you just picked the one factor you disagree with even if it has no measurable impact? Chronic under-financing of schools after all would disagree with the narrative that it is the kids fault.
Before grades were directly tied to school funding, a 9-month long school year included four semesters of varied learning, based on historically validated topics deemed as required learning in order to function and succeed in society as a graduating adult. It included one or two State standardized tests, just to ensure actual learning was taking place at the high school level designed to not merely create graduates, but prepare legal adults.
Today, grades are directly tied to school funding. Which change
Re: (Score:2)
Re: (Score:2)
The problem is not the goal. No Child Left Behind is perfectly reasonable. The issue is that there two different approaches to achieve this goal: you can lower the standards (cheap) or you can give stragglers the means to success (expensive). Guess what the typical politician will do.
Lowering the classroom standard to the lowest common denominator, does not create success. It creates a standard of mediocrity that races to the bottom as a goal.
Or you can take the intention of the project and focus on the word 'Left'. It shouldn't be about ensuring nobody is behind (impossible anyway) but about ensuring that nobody is LEFT behind purely because of their starting point. It's about not wasting talented students who just need a bit more extra attention to them for whatever reason. Not wasting resources is prudent conservatism.
Re: (Score:2)
My point stands. And I'm not confusing anything. The material covered in a course, and the level at which it is covered, are set before students enroll in the course. The pass/fail rate is determined by the kinds of students who enroll..
What, you mean human students? The “kinds” of blood-based fallible meatsacks that justify the concept of a bell curve every time?
Your blindness to the obvious expectation driven by human reality of 7% of students statistically expecting to fail, is likely a side effect of failed concepts like No Child Left Behind.
Yes. Some of those humans will fail, because humans. They can earn that failing grade regardless of what “kind” of student you assume they are. Why even keep track of CS
Standards? What standards? (Score:1)
Standards? Lowering the standards to the lowest common dumbnominator in the elementary classroom, is exactly how No Child Left Behind became No Moron Left Behind.
It’s also how we ultimately created illiteracy statistics in fucking college.
It’s going to be downright stupid scary when internet outages become a factor in measuring IQ. It’ll be rather easy to tell those who supplement their intelligence with AI from those who chose to wholly create their “intelligence” with it.
Re: (Score:2)
Re:Standards, not gaussians!! (Score:2)
Yes! Far too many people just blame the "teacher" when students fail... The culture in the USA of blaming everybody but yourself has been growing significantly over the last few generations and it began with the gen X parents (I know teachers old enough to have perspective.)
That said, I see students making an honest effort using the bot to tutor them. I've seen students who tried to use online "tutoring" services before that. In both cases, when I had honest interactions with those students, I noticed their
#1 reason - kids used AI to get into Berkeley (Score:5, Interesting)
Re: #1 reason - kids used AI to get into Berkeley (Score:2)
Re: (Score:2)
Who grades? (Score:1)
Where the grades coming from? They're not a gospel, the grades are being decided by professors. So the professors grade lower, and then they're "observing" lower grades. Hypocrisy for media attention!
Re: (Score:2)
Maybe the profs decided to stop pretending that these students actually learned something.
Re: (Score:2)
Root Cause. (Score:4, Insightful)
Stop allowing them to cheat so fucking much in high school.
And why do they do that? Well..
Stop selling grades for school funding. It’s how the concept of a bell curve has become a joke, and how cheating became acceptable.
Those aren’t “students” in high school. They’re financial sponsors.
Those aren’t “students” in college. They’re high-valued customers.
Re: (Score:3)
Education is essentially infrastructure. For-profit infrastructure does not work.
Re: (Score:2)
The evidence 'ai' makes ... (Score:2)
... humans dumber. Lazy students fail.
Things may be more complicated (Score:4, Insightful)
Cheating is one thing, "learning" with a LLM without learning is a second, and needs attention because a LLM *can* teach well, but the current standard use does not enable this. But the third thing universities struggle with is, what employers want. There is a good chance that the current degree loses worth as employers will seek people who are efficient with all tools available and have less priority on people being good (but not as good) without the tools.
This means that universities possibly need to shift the focus and reconsider how much things need to be learned so you can tell the formula when someone wakes you in the middle of the night or not. Maybe the skill how to get the maximum out of a modern tool is more valuable than knowing half of it without the tool. You can apply all your moral and ethic ideas about "real" knowledge if you want, but in the end they possible don't pay good anymore. It's cool if you can run a marathon, but your employer pays you to use the car.
Re: (Score:2)
Cognitive surrender at work (Score:2)
The only way to learn things and to understand them is to do your own thinking. Smaller minds do not understand that and hence try to do it without.
I am all for letting these people fail permanently. They have no place in the field they are trying to get into anyways. All they can do is damage.
Software in 2026 == Mud Pies? (Score:1)
why (Score:2)
Why should 7% receive D's and F's? What if everyone totally masters the material? What if the distribution is such that 99% of the students are within 1 point of each other? Do the profs need to devise tests for material that has not been taught, just to be sure that 7% fail? These people are nuts.
Next... (Score:2)
it'll be "Every agent get's a trophy". You think the hairless monkeys act entitled, just wait!
Intern's opnion (Score:2)
STEM achievement requires frequent mental labor (Score:2)
No edu data from 2020-2030 should be trusted (Score:2)
Not that it can't be believed, but that it should not be taken as definitive reality rather than temporary aberration.
Covid showed the huge fragility of most systems, but perhaps most frail was the process of learning.
We are now at the peak of a time interval during which nearly all college students are people who had their education and experiential development terribly disrupted right when they would be learning must-have foundations in logic, mathematics, and rhetoric. I would bet that even if we made no
Re: (Score:2)