what worked for me teaching an undergrad course last year was to have
in-class exams weigh 90% of the total grade, but let them drop their lowest score
take-home work weigh 10% and be graded on completion (which i announced to the class, of course)
i was also diligent about posting solutions (sometimes before the due date --- it's a completion grade after all) and i let students know that if they wanted direct feedback they could bring their solutions to office hours
\
it ended up working pretty well. an added benefit was that my TAs didn't have to deal with the nightmare of grading 120 very poorly written homeworks every four weeks. my students also stopped obsessing about the grades they would receive on their homeworks and instead focused on learning the grades they would receive on their exams
however, at the k-12 level, it feels like a much harder problem to tackle. parental involvement is the only solution i can think of, and that's already kind of a nightmare (at least here in the us)
is the deniability you are referring to of the clanker-wankers (CW^[unrelated, but i miss when that channel had superhero shows. bring back legends of tomorrow]) themselves or the clanker-producers (e.g. sam altman)?
because i agree on the latter^[i.e., someone like altman would say "you're prompting it wrong" to skirt accountability or create an air of scientific/mathematical rigor], but i do see CWs saying stupid shit like "there is more to it than just writing a description"
edit: credit, it was @antifuchs who introduced the term to me here
edit2: sorry, my dumbass understands your point now (i think). if i wank clankers and someone tells me "that shit doesn't work," i can just respond "you must have been prompting it wrong". but, i do think the way many users of these tools are so sycophantic means it's also a genuine belief, and not just a way to escape responsibility. these people are fart sniffers, after all
people who talk about "prompting" like it's a skill would take a class^[read: watch a youtube tutorial] on tasseomancy because a coffee shop opened across the street
is anyone else fucking sick and tired of discord? it's one thing if it's gaming-related^[i guess. not really, fuck discord.], but when i'm at a repo for some non-gaming project and they say "ask for help in our discord server", i feel like i'm in a fever dream and i'm going to wake up and discover that the simulation i was in was managed by chatgpt
an unintended side effect of this is people who can't or don't want to verify their age going to less reputable sources. so even though it can be done in a "privacy-respecting fashion" (see, for example, soatok's post on this^[https://soatok.blog/2025/07/31/age-verification-doesnt-need-to-be-a-privacy-footgun/] ), it's still a bad idea.
additionally, in my opinion no one who wants to enact such a thing is doing it in good faith. it is a pretense towards an ulterior goal^[e.g. "steam porn games" → "this person's existence is inherently sexual" → "ban lgbtq content"]
weird fuck's post reads to me as the mistake of thinking web/js is uniquely capable of dynamic code loading
what is stopping a desktop or mobile client from running new/different code? the only solution im aware of (we're in halting problem territory here, probably, though grapheneos has "prevent DCL from storage/memory" toggles so idk) is to inspect the code to make sure it does what they say and then cryptographically sign it
you can save even more time by not doing the work at all
the output is more consistent than what an LLM shits out, too
Edit: serious note, even though you probably aren't worth anyone's time: you may be conflating the technology's actual use cases (as an accountability sink and to spread misinformation) with the intentions of its creators. and the real reason higher-ups are pushing this is because they're pliant dipshits that would eat dogfood if the bowl was labelled "FOMO". also they hate paying employees
this is going to take up the rest of my evening, now i'm reading about the Pólya conjecture for the first time
this rules 💗