4 ms·
I just think it's really funny that a third or so of the article is the author struggling to figure out why this would be useful to anyone. > scanned and analy
by activitypea 3y ago
I just think it's really funny that a third or so of the article is the author struggling to figure out why this would be useful to anyone.
> scanned and analyzed a whole bunch of books and would let you call up really useful data on books. [...] Frankly, all of that sounds amazing. And amazingly useful. Even more amazing is that he built it, and it worked. It would produce useful analysis of books.
> This is all quite interesting. It’s also the kind of thing that data scientists do on all kinds of work for useful purposes. Smith built Prosecraft into Shaxpir, again, making it a more useful tool.
Author's general illiteracy aside, he's really giving the game away here. I can't even think about the ethical implications of the project, because why would I care to count the number of adverbs and passive voice in all books ever, and why would you need a state of the art LLM-powered AI to do it?
- GaggiX 3y agoThe ethical implication of indexing books? Also where do you read the author of the website was using LLM?
- em-bee 3y agohe is not even using an LLM, which is kind of the point. the tool is thrown in and judged as an AI tool when it is just simple statistics that anyone who studied some math can build with a little bit of effort.
- steezeburger 3y agoImagine seeing the trends in books throughout certain time period, wars, etc. Or larger trends over the history of all written works. Or all kinds of other neat and useful information that can impact decisions we make today. Do you think all analysis of things you aren't personally interested in is useless?
- activitypea 3y agoI hate "did you even read my comment" replies, but again, you're pitching me on "trends throughout a time period", "larger trends throughout history" and "all kinds of other useful information that can impact decisions". I don't think a person without interest in literary analysis would make this comment. On the contrary, I'm specifically calling out building non-sensical tools with little practical uses because the author of the tool does not engage with literary works beyond the most superficial layer. And because I don't believe quantitive analysis of art is benificial to culture and society, but that's a broader subject.