Neonlight - Computer Music
https://music.youtube.com/watch?v=SReThv8LJN4
WordAll #721 completed in 34s
๐ฉ๐ฉ๐ฉ๐ฉ ๐ perfect game!
wordall.xyz
Wordle 992 4/6*
โฌ๐จโฌโฌโฌ
โฌโฌ๐จโฌโฌ
๐จโฌ๐จ๐จโฌ
๐ฉ๐ฉ๐ฉ๐ฉ๐ฉ
https://www.nytimes.com/games/wordle/index.html
#Worldle #776 4/6 (100%) (cheated)
๐ฉ๐ฉ๐ฉโฌโฌโ๏ธ
๐ฉ๐ฉ๐ฉโฌโฌโ๏ธ
๐ฉ๐ฉ๐ฉ๐ฉ๐จโฌ ๏ธ
๐ฉ๐ฉ๐ฉ๐ฉ๐ฉ๐
๐งญโญ๐ฉ๐ค๐ช
La palabra del dรญa #791 X/6 (haunch section. Tonto de culo)
โฌ๐จ๐ฉโฌโฌ
โฌ๐ฉ๐ฉโฌ๐ฉ
โฌ๐ฉ๐ฉโฌ๐ฉ
โฌ๐ฉ๐ฉ๐ฉ๐ฉ
โฌ๐ฉ๐ฉ๐ฉ๐ฉ
โฌ๐ฉ๐ฉ๐ฉ๐ฉ
Le Mot (@WordleFR) #788 4/6
โฌโฌ๐จโฌโฌ
โฌโฌโฌ๐ฉ๐ฉ
โฌ๐จโฌ๐ฉ๐ฉ
๐ฉ๐ฉ๐ฉ๐ฉ๐ฉ
Framed #727 (nope. I watched Renia Roja and I am considering reading the book. The English translation, which like subtitles won't be 1:1 because it would sound odd having idiomatic stuff directly translated. I was trying to look up a Spanish insult, and maybe it's obscure, which is something like "Es mรกs fea que un pie sin uรฑas" which I wrote down in 2019 and it was probably from an old, 19th or 18th century, book. I think searches, Google particularly, have gotten dumber. Ironically because of AI. If normal distribution applies to the quality of information generated by people and their search terms, then ~68% of LLM training data is going to be within 1 standard deviation from the average. Obviously nature isn't exact but a potential outcome of LLMs maybe a more generalised mediocrity and mediocre will replace crap. So an overall improvement but potentially more difficult for the ~32% either side of the ~68%. And certainly more difficult when the searches returned are a good guess most of the time but irrelevant in a minority of cases, particularly when it the search engine opaquely overrules or applies interpretation to the search terms, which seems to be how it works now, along with ghastly SEO and paid ranks, with the ranking of results. It's like "did you mean?" but on an industrial scale. I think spellcheckers improve spelling, I think AI will improve people in general, I don't think technology is inherently bad, but how that reversion to mediocrity is avoided in LLMs could be better questions, with parameters, maybe even Boolean logic, and an "AI off" mode. To avoid a mediocrity loop in general purpose AIs.)
๐ฅ ๐ฅ ๐ฅ ๐ฅ ๐ฅ ๐ฅ ๐ฅ