@mttaggart I get some friends saying that the same stuff was said about computers, internet, even the printing press. But AI feels quite different to me. I resent the unpaid plunder of people's work for training data for machines that belong to billionaires. That has strong echos of colonial appropriation of art and knowledge, except we're all now in the colony being appropriated by a small class of essentially stateless billionaires. 1/2
Discussion
Loading...
@lindawoodrow
> I get some friends saying that the same stuff was said about computers, internet, even the printing press. But AI feels quite different to me
There is a rhetorical bait-and-switch here, of conflating #MOLE Training with AI in general, ie any use of automation or statistical methods to make computers seem "smart". It's similar to one of conflating corporate DataFarming platforms with the net in general, as if regulating the former is a threat to the latter.
@strypey I tried to search the internet for whether MOLE is a common derogatory term for certain LLM uses, but all I found was different machine learning models for identifying moles on your skin, and one article about a form of machine learning called motion learning. 🙂
@strypey Found your definition "Machine-Operated Learning Emulator" and your excellent article disintermedia.substack.com/p/i…
@clacke
> Found your definition "Machine-Operated Learning Emulator"
This was confusing people a lot, so I put a clarification in my profile, with a link to the article in its new home on my Ghost instance;
https://disintermedia.net.nz/invasion-of-the-mole-trainers/
Thanks for including the link to my post, I've edited it to swap out the SS link for that one.
> your excellent article
You're too kind ☺️
EDIT: 'edited', not 'exited', dang autocucumber : P