How does a machine reading a book fundamentally differ from a human
For this, we'd have to delve into the fringes of epistemology. What exactly does it mean "to read"? Is this the transfer of written word into concepts in the mind?
If so, then must not a machine be confirmed to have a mind before it is said to be reading rather than merely observing and using an input (text) to create an output (word probability map)?
Assuming no mind, does the act of a computerized camera observing text as an input equate to a copyright violation? If so, there are probably millions of copyright violations occurring daily right now, sans AI, even if no data is stored. If the output into a word probability map is the copyright violation, shouldn't use of the map be able to reproduce the original every time to be an encoded copy? But it's a probability map. It'll spew the same wonky BS your phone will spew unless the LLM stores metadata like probabilities of words after previous words/sentences/paragraphs in conjunction with the name of a particular work, making probabilities of 100%. But being an open system, the process should be available for all, and I bet things like names of works aren't intimately tied with the connections. Loosely at best, especially since the intent is to create a Large Language Model, and not a Large Library Model.