learning the deeply "fuck around and find out"-based history of organic chemistry makes a lot of what's happening with AI make sense to me
if massive compute at scale is "the new oil" might be good to spend more time learning about all the insane poisons and shit that were commercialized (with minimal research into health/environmental effects) to justify the waste production that went into refining oil for gasoline
@ingrid Any pointers for good reading on this? I’ve been pointing a lot of people lately at this on the early history and regulation of train safety for similar reasons, but I wonder if the “we can do it in a small lab” aspect makes chemistry even more apropos as ML gets faster/smaller/cheaper: https://www.hup.harvard.edu/catalog.php?isbn=9780674022614
@luis_in_brief i just finished Economic Poisoning by Adam Romero, which does not make the explicit connection to ML etc but is really rich re: industrial waste feeding into industrial ag https://www.ucpress.edu/book/9780520381568/economic-poisoning
@luis_in_brief @ingrid My friend Jon wrote this, which doesn’t talk about the ML parallels at all but was a great rundown of the environmental damage and how Standard Oil used things like regulatory capture and municipal “collaboration” to do it.
@luis_in_brief @ingrid (It was also just an easy read.)
@ingrid You speak like this isn't something we still do on a day-to-day basis...
@ingrid ⬆️
“Chemical Eye 👁️ on Knights Molecular” 👉 http://www.sitnews.us/MacDougall/062205_macdougall.html
Victorian gentlemen would just huff weird distillates they made in a lab and be like "huh, guess this is chloroform now" and sell it to dentists a couple of days later, AI has the same energy to me