31–40 of 601 posts
I simply don't waste my time reading an AD as an article. I take this as seriously as I would if McDonald's published articles about how much weight people lose eating at McDonald's.
Earlier quoted context omitted.
> We certainly improve productivity I think this is really still up for debate We produce more output certainly but if it's overall lower quality than previous output is that really "improved productivity"? There has to be a tipping point somewhere, where faster output of low quality work is actually decreasing productivity due to the efforts now required to keep the tower of garbage from toppling
It's not up for debate. Ask any programmer if LLMs improve productivity and the answer is 100% yes.
> A common question is: “how much are students using AI to cheat?” That’s hard to answer, especially as we don’t know the specific educational context where each of Claude’s responses is being used. I built a popular product that helps teachers with this problem. Yes, it's "hard to answer", but let's be honest... it's a very very widespread problem. I've talked to hundreds of teachers about this and it's a ubiquitous…
"The real danger lies in their seductive nature - over how tempting it becomes to immediately reach for the LLM to provide an answer, rather than taking a few moments to quietly ponder the problem on your own. By reaching for it to solve any problem at nearly an instinctual level you are completely failing to cultivate an intrinsically valuable skill - that of critical reasoning."
They were the first to adopt digital wordprocessing, presentations, printing and now generative AI even though in essence all of these would have been disproportionately more hand in glove for the humanities on a purely functional level.
It's just a matter of comfortability with and interest in technology.
> A common question is: “how much are students using AI to cheat?” That’s hard to answer, especially as we don’t know the specific educational context where each of Claude’s responses is being used. I built a popular product that helps teachers with this problem. Yes, it's "hard to answer", but let's be honest... it's a very very widespread problem. I've talked to hundreds of teachers about this and it's a ubiquitous…
Whenever we have a new technology there's a response "why do I need to learn X if I can always do Y", and more or less, it has proven true, although not immediately.
For instance, I'm not too concerned about my child's ability to write very legibly (most writing is done on computers), spell very well (spell check keeps us professional), reading a map to get around (GPS), etc
Not that these aren't noble things or worth doing, but they won't impact your life too much if you're not interest in penmanship, spelling, or cartography.
I believe LLMs are different (I am still stuck in the moral panic phase), but I think my children will have a different perspective (similar to how I feel about memorizing poetry and languages without garbage collection). So how do I answer my child when he asks "Why should I learn to do X if I can just ask an LLM and it will do it better than me"
> A common question is: “how much are students using AI to cheat?” That’s hard to answer, especially as we don’t know the specific educational context where each of Claude’s responses is being used. I built a popular product that helps teachers with this problem. Yes, it's "hard to answer", but let's be honest... it's a very very widespread problem. I've talked to hundreds of teachers about this and it's a ubiquitous…
> I think the issue is that it's so tempting to lean on AI. I remember long nights struggling to implement complex data structures in CS classes. I'd work on something for an hour before I'd have an epiphany and figure out what was wrong. But that struggling was ultimately necessary to really learn the concepts. With AI, I can simply copy/paste my code and say "hey, what's wrong with this code?" and it'll often spot…
I guess I'd qualify to you as someone "AI crutched" but I mostly use it for research and bouncing ideas (or code complete, which I've mentioned before - this is a great use of the tool and I wouldn't consider it a crutch, personally).
For instance, "parse this massive log output, and highlight anything interesting you see or any areas that may be a problem, and give me your theories."
Lots of times its wrong. Sometimes its right. Sometimes, its response gives me an idea that leads to another direction. It's essentially how I was using google + stack overflow ten years ago - see your list of answers, use your intuition, knowledge, and expertise to find the one most applicable to you, continue.
This "crutch" is essentially the same one I've always used, just in different form. I find it pretty good at doing code review for myself before I submit something more formal, to catch any embarrassing or glaringly obvious bugs or incorrect test cases. I would be wary of the dev that refused to use tools out of some principled stand like this, just as I'd be wary of a dev that overly relied on them. There is a balance.
Now, if all you know are these tools and the workflow you described, yea, that's probably detrimental to growth.
I simply don't waste my time reading an AD as an article. I take this as seriously as I would if McDonald's published articles about how much weight people lose eating at McDonald's.
It's more like an analysis of what items people order from McDonald's, using McDonald's own data which is otherwise very difficult to collect. Your loss!
Interesting article, but I think it downplays the incidence of students using Claude as an alternative to building foundational skills. I could easily see conversations that they outline as "Collaborative" primarily being a user walking Claude through multi-part problems or asking it to produce justifications for answers that students add to assignments.
> Interesting article, but I think it downplays the incidence of students using Claude as an alternative to building foundational skills. No shit. This is anecdotal evidence, but I was recently teaching a university CS class as a guest lecturer (at a somewhat below-average university), and almost all the students were basically copy-pasting task descriptions and error messages into ChatGPT in lieu of actually program…
People who spent the past two years offloading their entry-level work onto LLMs are now taking 400-level systems programming courses and running face-first into a capability wall. I try my best to help, but there's only so much I can do when basic concepts like structs and pointer manipulation get blank stares.
> "Oh, the foo field in that struct should be signed instead of unsigned."
> "Yeah, the type definition of Bar? It's right there."
> "..."
I feel like Anthropic has an incentive to minimize how much students use LLMs to write their papers for them. In the article, I guess this would be buried in > Students also frequently used Claude to provide technical explanations or solutions for academic assignments (33.5%)—working with AI to debug and fix errors in coding assignments, implement programming algorithms and data structures, and explain or solve mathe…
Most of their categories have straightforward interpretations in terms of students using the tool to cheat. They don't seem to want to/care to analyze that further and determine which are really cheating and which are more productive uses. I think that's a bit telling on their motivations (esp. given their recent large institutional deals with universities).
> We found that students primarily use Claude to create and improve educational content across disciplines (39.3% of conversations). This often entailed designing practice questions, editing essays, or summarizing academic material.
Sure, throwing a paragraph of an essay at Claude and asking it to turn it into a 3-page essay could have been categorized as "editing" the essay.
And it seems pretty naked the way they lump "editing an essay" in with "designing practice questions," which are clearly very different uses, even in the most generous interpretation.
I'm not saying that the vast majority of students do use AI to cheat, but I do want to say that, if they did, you could probably write this exact same article and tell no lies, and simply sweep all the cheating under titles like "create and improve educational content."