blakestacey @ blakestacey @awful.systems

Posts

52

Comments

1,470

Joined

2 yr. ago

Moderating

---

A belief system that inculates the believer into thinking that the work is the most important duty a human can perform, while also isolating them behind impenetrable pseudo-intellectual esoterica, while also funneling them into economic precarity... sounds like a recipe for delicious brownies trouble.

From Lila Byock:

A 4th grader was assigned to design a book cover for Pippi Longstocking using Adobe for Education.

The result is, in technical terms, four pictures of a schoolgirl waifu in fetishwear.

Dr. Casey Fiesler reports,

I was poking around Google Scholar for publications about the relationship between chatbots and wellness. Oh how useful: a systematic literature review! Let's dig into the findings.

[...]

Did you guess "that paper does not actually exist"?

Did you also guess that NOT A SINGLE PAPER IN THEIR REFERENCES APPEARS TO EXIST? [...] When I was searching in various places to confirm that those citations were fabricated, Google's AI overview just kept the con going.

Jill Walker Rettberg in the comments:

There’s a peer reviewed published paper in AI & Society called Cognitive Imperialism and Artificial Intelligence which is clearly mostly AI-generated. Citations are real but almost all irrelevant. I emailed the editors weeks ago but it’s still up there and getting cited.

(thinks) groxxing

This chatbot-generated house floor plan gave me a 45-second giggle fit.

(The original post does seem to be satirical in intent.)

At least this example grew out of actual humans being suspicious.

Dozens of academics have raised concerns on social media about manuscripts and peer reviews submitted to the organizers of next year’s International Conference on Learning Representations (ICLR), an annual gathering of specialists in machine learning. Among other things, they flagged hallucinated citations and suspiciously long and vague feedback on their work.

Graham Neubig, an AI researcher at Carnegie Mellon University in Pittsburgh, Pennsylvania, was one of those who received peer reviews that seemed to have been produced using large language models (LLMs). The reports, he says, were “very verbose with lots of bullet points” and requested analyses that were not “the standard statistical analyses that reviewers ask for in typical AI or machine-learning papers.”

We seem to be in a situation where everybody knows that the review process has broken down, but the "studies" that show it are criti-hype.

Welcome to the abyss. It sucks here (academic edition).

I used https://tektite.cc/ to migrate off Bluesky, and picked the myatproto.social option from the drop-down list. This may be a good start: https://leaflet.pub/000b57de-78dc-4939-8c66-79227d010cce

Chasing links landed me here:

Grimes used to come to my club nights in Vancouver, and one time a guy who didn't know who she was saw her dancing like an attention starved, crystal-gripping idiot, and he said to me "That's the kind of chick who would take a shit on your chest if you asked."

https://blacksky.community/profile/did:plc:mpc62tgblkwndximirue5dxg/post/3lh7kznna3k2y

T Kamal in the comments:

I mean, not to tar any groups with a broad brush, but like, if I had to be criticized by any one person, I'd prefer to be criticized by DAIR rather than EAs, because the EAs hang around the Rationalists, and the Rationalists birthed the Zizians, and those folks have murders attributed to them. Like, at least with Dr. Gebru I can be reasonably sure that no one's gonna come at me with a katana.

Noted on Bluesky:

Tomorrow Grimes will DJ a livestream of immortality influencer Bryan Johnson tripping on shrooms to determine its effect on longevity. Mr. Beast and the CEO of Salesforce will be there too.

Now, folks out there are calling this a Biblically accurate blunt rotation, but to be fair, it's missing Aella.

Years ago, I said, "I've never finished a Stephenson novel." Someone replied, "Neither has he."

Scientific Reports did not have what one would call a sterling reputation prior to this. Mathematical physicist John Baez wrote,

If you're a physics crackpot who wants to publish in a prestigious-sounding journal, I recommend Nature Scientific Reports! You have a good chance of getting your paper in!

Try making it look like "Mass–Energy Equivalence Extension onto a Superfluid Quantum Vacuum". [...] This paper looks like a lot of the emails I get. It would never be published in a serious physics journal:

https://www.nature.com/articles/s41598-019-48018-2

And it's not the only crackpot physics paper that's been published by Nature Scientific Reports!

Here's a much crazier paper in Nature Scientific Reports:

https://nature.com/articles/s41598-019-46765-w

It's called Maximum Entropy (Most Likely) Double Helical and Double Logarithmic Spiral Trajectories in Space-Time. You have to read it!

My guess is that Nature Scientific Reports doesn't have mechanisms built in to enforce the oppressive hidebound orthodoxy that dominates the other physics journals. So if you have a revolutionary new theory, submit your paper here!!!

Flavio Nogueira in the comments:

I have been in a meeting with the editors of SRs and its editor in chief years ago during an APS March Meeting. I can tell you that some editors were truly pissed off, as papers rejected after peer reviewing ended up being published anyway...

What word do they want to unambiguously represent that concept now?

"Literally, not figuratively", said in a Sterling Archer voice.

The use of literally in a fashion that is hyperbolic or metaphoric is not new—evidence of this use dates back to 1769. Its inclusion in a dictionary isn't new either; the entry for literally in our 1909 unabridged dictionary states that the word is “often used hyperbolically; as, he literally flew.”

— Merriam-Webster

Out of all the documentation pages written by the most pedantic nerds on Earth (complimentary), I think this just might be the shortest.

https://en.wikipedia.org/wiki/Wikipedia:Writing_articles_with_large_language_models

Not all, but more than half (at ICLR).

There's a PubPeer thread.

Good day to remember the rat dck pck.

And this is not the first time Scientific Reports has been sneered.

... Autons?

Tag yourself, I'm "runctitional features"

I've been nursing a grudge against New York Magazine for fifteen years.