skisnow

joined 2 months ago
[–] skisnow@lemmy.ca 7 points 1 day ago

Yeah, the second one. It's the ones prepared to do shit like that who get promoted in the first place.

[–] skisnow@lemmy.ca 7 points 1 day ago (1 children)

This is true, but it’s also not helped by the fact that the web itself has gotten worse. A lot of stuff that previously would have been on a crawlable page is now on an amorphous infinite scroll feed behind a login screen, or worse still a YouTube video.

[–] skisnow@lemmy.ca 4 points 2 days ago

Good job there’s more than 20 other studies out there for them to p-hack

[–] skisnow@lemmy.ca 3 points 2 days ago* (last edited 2 days ago) (1 children)

Having spent some small time in the information theory and signal processing world, it infuriates me how often people champion LLMs for writing things like data dictionaries and documentation.

Information is measured in information theory as "the difference between what you expected and what you got", ergo, any documentation generated automatically by an LLM is by definition free of Information. If you want something explained to you in English then it can be generated just as easily as and when you want it, rather than stored as the authoritative record.

[–] skisnow@lemmy.ca 2 points 2 days ago

What's currently pickling my noggin is how I've been seeing "new model smashes benchmarks by an unexpectedly huge factor" headlines every month for the last two years, and yet somehow no matter how many models suddenly score 99% on tasks that they used to score 20% for, I've not actually found the damn thing any more helpful or reliable than it was in 2023 for anything real-world. I'm starting to think all these supposed breakthroughs they keep having are being hugely overstated.

[–] skisnow@lemmy.ca 12 points 2 days ago

JFC that fucking pisspoor AI photo they used. I fucking hate the future so much.

[–] skisnow@lemmy.ca 13 points 2 days ago* (last edited 2 days ago)

Yeah, kind of my point was that y'all are viewing it as a "field trip", which is typically a specific event that's infrequent, carefully organized and supervised, which is a whole different beast to the generic standing instructions of "we're not going to supervise your kids if they wander off school grounds" slip.

For the former case it's pretty much understood that everyone in class should be able to join a field trip, but for the latter it's not unusual for parents to decline and therefore teachers would absolutely be expected to enforce the rules.

[–] skisnow@lemmy.ca 47 points 2 days ago (3 children)

For the benefit of the many non-Brits complaining about how unrealistic it is: the Leaving School Grounds Unsupervised form is (when I grew up at least) a huge social divider and Big Deal in a lot of British schools. There was a whole micro industry at mine where the ~70% of kids who were allowed out would provide delivery services for sweets and pop for the 30% who weren’t.

JKR didn’t just pull this whole thing out her ass, it was something that most British kids will have instantly related to. (She’s still an awful human mind)

[–] skisnow@lemmy.ca 7 points 2 days ago* (last edited 2 days ago) (3 children)

The “socialists expect people to do stuff for free” trope only exists in capitalist strawman rhetoric.

[–] skisnow@lemmy.ca 1 points 3 days ago

All bases should be quoted in Base-1, the basedest of bases

[–] skisnow@lemmy.ca 8 points 3 days ago* (last edited 3 days ago)

One of America's many diseases is its staunch unwillingness to send company executives to jail for doing crimes, despite otherwise having the highest incarceration rate worldwide.

view more: ‹ prev next ›