Miranda Osthoff
banner
tinyconstruct.bsky.social
Miranda Osthoff
@tinyconstruct.bsky.social
MPLS computer witch
Indie TTRPG goblin
Books, dog person
Tired 🏳️‍🌈🏳️‍⚧️
They hate to see other mediocre people being held accountable
December 2, 2025 at 8:04 AM
You’re bad at reading psychology essays then. The rubric is included in all kinds of articles, the TA and professor both have her a zero, AND she was being an inflammatory, bigoted shit, clearly on purpose, on top of writing an essay that earned 0 points.
December 2, 2025 at 8:03 AM
Seems a fair price for reduced muscle mass, depression, kidney dysfunction, loss of vision, GI issues, and colon malfunction in exchange for looking weird and not being healthier.
December 2, 2025 at 7:52 AM
“I support trans rights but [I think trans people should live in fear of retributive cis people, who always are in the right, no matter their difference in professional expertise or moral goodness]” what dogshit.
December 2, 2025 at 7:36 AM
Hey maybe a professor of psychology who just won an award in graduate teaching knows more about grading papers than you. In fact, having *read* the essay, how in the fuck can anyone think it would merit a passing grade for a *high school* student?
December 2, 2025 at 7:34 AM
Could you possibly do a single thing that matters

bsky.app/profile/jcha...
This evening, a resident was kidnapped by agents with face coverings. They disguised themselves in masks and as Uber drivers. They even left kids abandoned inside a car in the freezing cold. 1/
December 2, 2025 at 7:14 AM
I'm...honestly going to take more trains. At this point they are like the same price instead of more expensive, sure they are slower but guess what airlines get delayed constantly now and have become less safe, I don't get treated like shit when I use them, less carbon impact
December 1, 2025 at 7:16 PM
Yuuuup or therapy (provably more effective than SSRIs) would just be normal healthcare, also we’d have public healthcare, also we’d make life idk affordable at all and rates of depression and drug use would fall 🫠🫠🫠
December 1, 2025 at 3:33 PM
I tried like 5 different meds for depression, none of which did anything; Vyvanse made a huge difference
December 1, 2025 at 3:26 PM
“The name of your academic institution” hmm definitely not part of their considerations I’m sure
Anyway bye forever
November 30, 2025 at 10:20 PM
Which is a system that lets them *not* give credits to people doing research that they suspect will show how not-useful AI is. So, thank you for the thought exercise that has convinced me this is even more obnoxious than the way I was thinking of it initially.
November 30, 2025 at 10:13 PM
Also the “free credits for research” part is not guaranteed; Anthropic doesn’t just give those out to whoever, there’s an application process. So the difference between this and “We received a grant from Pfizer equal to or in excess of material costs” is there isn’t one
November 30, 2025 at 10:11 PM
But beyond the study, the bigger issue IMO is that by providing a product (that is proven to not be very effective) for free they are making CS students/departments dependent on it, also provably dumber.
November 30, 2025 at 10:06 PM
It is absolutely a source of potential bias. Those credits translate into literal power/maintenance/product costs. They do not state what *kind* of credits are provided, which I also take small issue with. Literally “have this thing worth money, from us, a corporation.
November 30, 2025 at 10:05 PM
Turns out you need things that cost money…to do studies. Do I think it noteworthy that despite this potential bias they reported such a low degree of usefulness? lol yes
November 30, 2025 at 9:59 PM
Do I think providing material that directly translates to money counts as funding? Yeah
November 30, 2025 at 9:58 PM
Did you read the paper, which pretty clearly says they were given API credits to do the study?

iceberg.mit.edu/report.pdf
November 30, 2025 at 9:48 PM
Me when my players prefer D&D and maps over OSRs and locations they can travel to because it’s “more choice-driven”
November 30, 2025 at 9:33 PM
To be clear, people are talking about LLMs in generative AI. What are you going to use those for in a way that is publicly ethical, and in a way that doesn’t require massive compute farms (because they are inherently unethical)
November 30, 2025 at 7:11 PM
(And even then, when my partner started playing 2014 5e their reaction was “Wow, there’s some racism cooked into this game.”)
November 30, 2025 at 6:59 PM
Even if someone is cynical and doesn’t care about all the real moral failings of WotC, 5e asking “Hey what if this wasn’t unpalatable to everyone except cishet white men” was a good financial move, folks. Wild how that works.
November 30, 2025 at 6:58 PM
The justification of “X people are bad for Y reason, I won’t support this” always comes out, but the reality of “I am afraid of being criticized for hating people, which I will not be examining” is loudly unspoken
November 30, 2025 at 6:16 PM