Remote
system asta
@ah@goto.boserup.eu
I literally do not see replies from randos. I do not approve follow requests from big-instance dwellers with empty bios.
Elden Ring, LLM Kompromat, Emacs, Life, Modal Editing. I don't know anymore. I post in Danish and English with the occasional Brecht lyrics in German.
0 Followers
0 Following
23 Posts
Joined April 21, 2025
pronouns:
she/her
homepage:
kompromat:
Replying to
@simon_jf I recently had the pleasure of
- "Are you here to see the guest lecture?"
- "I'm giving the guest lecture."
Replying to
@folfdk Ja. Jeg er nysgerrig på om den på de 3.000 mest gav falskt positive eller falskt negative eller hip-som-hap. Hvis man bliver clearet af maskinlæringen, hvad er så sandsynligheden for at en radiolog ville gøre det samme? Det er lidt trælst at blive kaldt tilbage dagen efter fordi nu har et menneske endelig kigget på det.
Replying to
@folfdk Enig. Jeg tænker også falsk negative er det værste. Jeg ville i hvert fald hellere vente lidt længere på at blive clearet af en radiolog når jeg alligevel er der, end sove på en brækket arm og skulle rakke afsted igen dagen efter. Jeg tænker også statistik og den slags bliver et vigtigere fag i skolen!
Og, ja, vi er ikke ude i chatbots her, men god gammel machine learning. Jeg mindes en model for nogle år siden der bare clearede alle dem hvor røntgenbilledet ikke havde navn på og flaggede dem med skrift nede i hjørnet. Det var nemlig den opdeling træningsdataen havde haft. Selve røntgenbilledet var ligegyldigt for konklusionen. Jeg tænker de er blevet bedre i dag.
Replying to
@tantramar I feel like one's client could cache the last couple of days of timeline for this kind of search. If it doesn't have to go too far back, then it seems like a reasonable client thing.
Lose my job to AI? I will simply lose my job to the passing of time. Get left behind.
A very first post on the Effekt blog:
https://effekt-lang.org/blog/recursion-schemes/
Replying to
@liamoc My old physics teacher used to say something like "the more you can struggle, the faster you can learn.". I also don't have any good solutions, but I sympathize.
Replying to
@datarama ouch! Gute Besserung! I can't help but laugh at the idea of sleep-lifting, though!
You! Yes, you! You should submit a paper to Dalí 2025, the 6th Workshop on Dynamic Logic - New trends and applications.
The PC is chock-full of great people, so send in a short or regular paper on anything to do with dynamic logic!
http://www.dali2025.cn/
Replying to
@datarama@hachyderm.io damn I have a ninth tooth, but I don't have however many this picture has!
Replying to
@datarama@hachyderm.io seems like a pretty good system, actually. If one insists on snacking jellyfish.