Dort_Owl [they/them, any]@hexbear.net to technology@hexbear.netEnglish · 13 days agoAI Fiction Is Easy to Detect Because It's Stupid and Bad, Research Findswww.404media.coexternal-linkmessage-square35linkfedilinkarrow-up198arrow-down10
arrow-up198arrow-down1external-linkAI Fiction Is Easy to Detect Because It's Stupid and Bad, Research Findswww.404media.coDort_Owl [they/them, any]@hexbear.net to technology@hexbear.netEnglish · 13 days agomessage-square35linkfedilink
minus-squareKuroXppi [they/them]@hexbear.netlinkfedilinkEnglisharrow-up21·13 days agoWaa a study written in partnership with Google DeepMind and the author used AI to implement her code and edit her writing. "I am a big believer that AI can help or hurt writing, but usually helps when not used to create more internet ‘slop’.” Cool
minus-squarechgxvjh [he/him, comrade/them]@hexbear.netlinkfedilinkEnglisharrow-up12·12 days agoI see that so much on forums recently Me: Why are you posting slop OP? OP: I’m just using LLMs for spell checking. The post: 1000-2000 words structured with headings (usually more or less the same headings) and m dashes for what is usually a 100-200 word question. I’ve also seen someone proclaim that they ran their slop through a deslopify slop tool to make it less sloppy.
minus-squarejohn_brown [comrade/them]@hexbear.netlinkfedilinkEnglisharrow-up5·12 days agoI regularly get clients pasting the entire LLM output into emails, including the preamble where the LLM tells the client how good its output is
Waa a study written in partnership with Google DeepMind and the author used AI to implement her code and edit her writing.
Cool
I see that so much on forums recently
Me: Why are you posting slop OP?
OP: I’m just using LLMs for spell checking.
The post: 1000-2000 words structured with headings (usually more or less the same headings) and m dashes for what is usually a 100-200 word question.
I’ve also seen someone proclaim that they ran their slop through a deslopify slop tool to make it less sloppy.
I regularly get clients pasting the entire LLM output into emails, including the preamble where the LLM tells the client how good its output is