Vibeschism aggregates links regarding LLMs from Hacker News, then I myself, a human, will review the link and score it -2 through 2.
Note that only HN scores >= 200 are factored.
Note that scoring is completely and totally subjective, based on my personal opinion regarding a link's sentiment. Also note that an LLM was indeed used to generate parts of the web app itself. (Even though I had to berate it constantly for doing assine things.)
Note that it's completely possible that articles which did not contain explicit LLM terms may have fallen through the cracks, e.g. "Vouch".
SENTIMENT SCORING A link is given a PREDICTION SCORE and a DISPOSITION SCORE, then that score is multiplied by the amount of points the link got on HN. PREDICTION bull 2 (explicitly) the future will have more LLMs bull 1 (implicitly) the future will have more LLMs, "bull" growth if only by virtue of "more" and "captured attention", might describe any positive aspects of LLMs 0 no prediction or prediction cancels/balances out bear -1 LLM usage will flatten out or stagnate (avoiding LLMs, prefers coding by hand, scalability concerns), may fight LLMs in some capacity, LLM abuse, legal concerns bear -2 LLMs are a fad, humans will always need to be in the loop, the bubble is going to pop DISPOSITION love 2 pushing people into using LLMs, drumming up hype, tone may be scornful of non-assisted coding, alack of substance, thinly veiled ad, or meth addled grandiosity love 1 LLMs are useful, tutorials, press releases, new models, tone is metered and respectful 0 ambivalent, remains aware of LLMS, cautious or no opinion, cancels/balances out hate -1 wary of the future that is in front of us, distaste of LLM output/companies, suggests use cases that LLMs do poorly hate -2 contempt, somewhere in the 5 stages of grief. describes malfesance, corruption, or fraud. describes banning or serverely limiting LLM usage