Cyberpunk Librarian

A guy named Vess once explained Bitcoin and crypto as "It's like if idling your car 24/7 occasionally produced solved Sudoku puzzles that you could then exchange for heroin."

AI is like restarting a failed nuclear reactor to produce enough energy to occasionally offer a semi-correct answer to a complex question while simultaneously being unable to identify that there are three r's in the word 'strawberry.'

Dec 30, 2024 at 4:57pmWeb
Show threadCassiDec 30, 2024
@CyberpunkLibrarian The annoying thing is that companies are using AI to screen candidates, and the only way to remain competitive is to tailor your resume as much as possible to get a high match score that may or may not cause an actual human to look at your application. Given the high competition, the more applications, the better, and it's very time-consuming to do all the keyword-checking manually. I really hate that it's come to this.
Show threadCyberpunk LibrarianDec 30, 2024

@Cassiopeia12727 My wife works in HR and she's not keen on all the AI nonsense either. If for no other reason than you really want to check the AI's work if you're looking for candidates with specific skillsets and abilities.

Well, if you need to check the AI's work... why have it at all? It's like, yeah, we used to have to sort through a pile of applications. Now we let the AI do that, and then we sort through the pile anyway to make sure it didn't miss someone good.

Progress?

Show threadCassiDec 30, 2024
@CyberpunkLibrarian I'm glad to hear that there are people in HR who are opposed to the usage of AI in the recruiting process who take time to audit the results. I'm sure it's hard to convince an organization to not use AI for HR.
Show threadpaulDec 31, 2024
@CyberpunkLibrarian
Show threadStarkRGDec 31, 2024
@amiserabilist @CyberpunkLibrarian I told my stepdad that crypto isn't as much of a scam as NFTs, but it's still mostly a scam, and an extremely environmentally unfriendly one at that.
Show threadTrump & Epstein: Biz partners.Jan 3, 2025
@StarkRG @amiserabilist @CyberpunkLibrarian
And naturally, The #ElongatedMuskrat... who is a HUGE #Dogecoin #CryptoBro... has convinced #DonTheCon (who once ridiculed #Crypto) to start one of his own.
https://www.thenation.com/article/economy/trump-federal-crypto-reserve/
Show threadDavid BruchmannJan 1, 2025

@amiserabilist @CyberpunkLibrarian

Distinct vocabulary is so important! 😂

Show threadDave NearyJan 3, 2025
@amiserabilist @CyberpunkLibrarian I love this! There are so many ways that the wolves unburden the sheep, it's a shame to call them all Ponzi schemes!
Show threadHavyhh2Jan 3, 2025
@amiserabilist @CyberpunkLibrarian ...though the word "scheme" seems a prominent element of these.....schemes...
Show threadferricoxideDec 31, 2024
@CyberpunkLibrarian

Having grown up a couple miles outside the TMI evacuation-perimeter, it boggles my mind that they're going to bring it back online to power AI.
Show threadCyberpunk LibrarianDec 31, 2024

@ferricoxide No no no! You don't understand! They're not reactivating Three Mile Island! No, they're powering up...

checks notes

Crane Clean Energy Center.

See there? That's much more friendly, right?

Show threadferricoxideDec 31, 2024
@CyberpunkLibrarian

#Lulz
Show threadRichard W. Woodley ELBOWS UP 🇨🇦🌹🚴‍♂️📷 🗺️Dec 31, 2024

@CyberpunkLibrarian

apt metaphors

Show threadSteven G. HarmsDec 31, 2024

@CyberpunkLibrarian

I get the ai skepticism too, but I had an interesting discussion about Neoplatonic elements in an anime I was watching (with Claude) and the takes were as good as anyone I might have had junior seminars with in my philosophy degree.

For “r” in strawberry, no; for write this tricky code error free in Python 3.12 maybe not; but for contextual discussions it’s not 0-value.

Arguably a more #xanadu or better linked data internet would bury that value per unit energy cost, but…we don’t have those.

Show threadCP93Dec 31, 2024
@CyberpunkLibrarian
AI: world’s worst autocomplete.
Show threadΈλλεν Εμίλια Ά.ζ.Jan 2, 2025
@CP93 @CyberpunkLibrarian world's most expensive autocomplete
Show threadJust Another AmyJan 1, 2025
@CyberpunkLibrarian
Large language models — which is what I assume you mean by “AI” here — are frequently used to retrieve, aggregate, and manage information, but that’s not what they’re *for*.
That’s a misuse to justify hype and funding-seeking.
They identify and replicate patterns in naturalistic language. That’s it.
If you want to mock them for something… they’re high powered predictive text that people *think* can give them answers.
Show threadDragon-sided DJan 2, 2025

@justanotheramy @CyberpunkLibrarian

I'm not sure whether your definition includes AlphaFold