Post Ax4NSZ7T28DlAIDOG8 by picofarad@noauthority.social
 (DIR) More posts by picofarad@noauthority.social
 (DIR) Post #Ax4NCcOd5BIoh7ddCa by davidtoddmccarty@me.dm
       2025-08-11T18:52:41Z
       
       0 likes, 0 repeats
       
       Do we think that GenAI is going to make it over the hump at some point and prove its usefulness? I wouldn’t bet anything on it myself. I don’t care if they get it to work, which is sort of the point for me. I don’t want it.
       
 (DIR) Post #Ax4NCdhSEfxGjnSBBQ by picofarad@noauthority.social
       2025-08-11T19:15:32Z
       
       0 likes, 0 repeats
       
       @davidtoddmccarty it depends; are humans more than biological algorithms that they don't understand?For instance, if all you try to model is the human brain, and no externalities, i think the project will fail to be "gen AI". In the sort of way that some towns have peculiarities, that, when asked about, are explained away as "we've always done it that way", vestigial reminders of the past.There's a lot of evidence that humans are more gut than brain. Failing to account for this in "AI" ...
       
 (DIR) Post #Ax4NSZ7T28DlAIDOG8 by picofarad@noauthority.social
       2025-08-11T19:18:27Z
       
       0 likes, 0 repeats
       
       @davidtoddmccarty as it stands *right now*, all AI is equivalent to "most likely" answers/images to "prompts" based on the training corpus; that is, it's akin to full access to the largest library, one that has all historical periodicals from 2005 or so, encyclopedias, dictionaries, and hard copies of online forums.A deft librarian can pull a cite up quite fast, but we don't say the librarian "knows" what the content is. generally.