3 ms·
Does this sound old-fashioned? Read more books and less Internet. There is a lot of information in books that is not on the Internet, yet many people don’t see
by chubot 4y ago
Does this sound old-fashioned? Read more books and less Internet.
There is a lot of information in books that is not on the Internet, yet many people don’t seem to realize it. I notice a lot of people seem to get almost all of their information from the Internet. Over time, I think that will make you stupid.
(The neutering of google books was a setback, but perhaps paradoxically good — it preserved a unique value of books)
There are also a lot of fluffy books published in the last 20 or so years - roughly equivalent to content farms - which is why older books that have stood the test of time are a good bet.
- seydor 4y agoDon't they use all the out-of-copyright books in the training ?
- chubot 4y agoYeah that's a good question. Google does serve all the out-of-copyright books, but I know nothing about the inputs to the LLMs. I think they still have big computational limitations, i.e. it doesn't "just scale" like web crawling and indexing. There's probably a way to probabilistically test from the outside if a given work is in the input set, or at least if some work is