Conversation

Notices

1. feld (feld@bikeshed.party)'s status on Friday, 05-Jan-2024 03:08:24 JST feld

   • SlicerDicer
   > AI and LLMs and enshittification is going to make the internet worse
   
   No, it's going to make it better. It was already terrible. I was just talking to @SlicerDicer the other day about the demise of the American Chestnut. It was a terrible blight and there are extremely few resistant trees in existence, but the timelines were all wrong on the search results we found. Wikipedia was wrong about it (although another article on Wikipedia was mostly right; it's annoying when 2 articles on Wikipedia cover nearly the same information and significantly diverge)
   
   so @SlicerDicer pulled out his book "Seeds of Woody Plants in the United States" written by the US Dept of Agriculture / Forest Service from 1974 which was a revision of the previous book in 1948. The book documented the official timeline when the blight was discovered in 1904.
   
   The internet has always been shit at telling the whole truth.
   • Disinformation Purveyor :verified_think: likes this.
   • SlicerDicer (slicerdicer@bikeshed.party)'s status on Friday, 05-Jan-2024 03:08:23 JST SlicerDicer

     in reply to
     @feld Having the old data written by trusted sources makes major differences.
     Disinformation Purveyor :verified_think: likes this.
   • Disinformation Purveyor :verified_think: (thatguyoverthere@shitposter.club)'s status on Friday, 05-Jan-2024 03:10:20 JST Disinformation Purveyor :verified_think:

     in reply to

     • SlicerDicer
     @feld @SlicerDicer but if llms are just being fed the bad data (even if they are fed the good data it seems like the good to bad ratio is heavily skewed in one direction) how can anyone expect them to produce more truthful information/improve the situation?