Can the Internet Archive claim that it’s developing it’s own AI and should have rights to scan everything and serve it to “customers?”
You can read the IA’s take on these issues here: https://blog.archive.org/2023/11/02/internet-archive-submits-comments-on-copyright-and-artificial-intelligence/
Their take is quite different from the sentiment that seems to prevail in this community.
Piracy for me but not for thee!
What I don’t see is an explanation for why developing AI should be a priority. I don’t even think LLMs are really AI, so why should they get free poor plagiarism? I haven’t seen any convincing evidence that LLMs are a step towards AGI.
I am really dreading the LLM crime prediction future. It will be like a really bad version of Minority Report. CRIMESTOPPER 3000 says you’re a minority, you’re poor, and you criticize our largest shareholder, therefore you are going to jail to prevent your inevitable crime.
We should be powering these energy hogs off before they make the majority of us poorer.
Honest question. If develop an “AI” that has perfect memory and just repeats verbatim what has been trained on… Can i just download books and distrubute them with that?
