plinky [he/him]@hexbear.net to technology@hexbear.netEnglish · edit-22 months agoGeneralization bias in large language model summarization of scientific researchhexbear.netimagemessage-square7linkfedilinkarrow-up158arrow-down10file-text
arrow-up158arrow-down1imageGeneralization bias in large language model summarization of scientific researchhexbear.netplinky [he/him]@hexbear.net to technology@hexbear.netEnglish · edit-22 months agomessage-square7linkfedilinkfile-text
linky to article table 3 contains examples of mutations, some of them seem fairly benign, but anyway don't blindly trust chatbot summaries
minus-squareD61 [any]@hexbear.netlinkfedilinkEnglisharrow-up9·edit-22 months agoCouldn’t we just hire a fuckton of people whose sole job would be to read research papers and work to create “normie” readable abstracts?
minus-squarePhilosoraptor [he/him, comrade/them]@hexbear.netlinkfedilinkEnglisharrow-up11·2 months agoThat’s what a science writer is supposed to be.
minus-squaretocopherol [any]@hexbear.netlinkfedilinkEnglisharrow-up6·2 months agoWe had that, various popular science journals, Nature articles, decent newspapers would be able to do that, but it’s not profitable unless they are paid to shill for a corporation.
Couldn’t we just hire a fuckton of people whose sole job would be to read research papers and work to create “normie” readable abstracts?
That’s what a science writer is supposed to be.
We had that, various popular science journals, Nature articles, decent newspapers would be able to do that, but it’s not profitable unless they are paid to shill for a corporation.