"All translations in this collection are © Scott Weisman. All rights reserved, except as granted by the license below."<p>Does copyright actually belong to Scott Weisman if all the words are outputs of LLMs? Is there any relevant case law here? I'm very curious about whether LLM outputs are copyrightable by the LLM user (I'm guessing potentially it varies throughout the world).
Solely the direct output of an LLM? Not in the US, no. Can a human edited and corrected text in which a human adds authorship (corrections, style, diction, editing, composition etc) to LLM output? Yes. Though simply making singular edits/corrections probably wouldn’t meet that bar; this seems like a lot more than that.<p>An original human-authored translation is a derivative work that can hold copyright, but it is the human-authored parts that give it this status and protection.<p>The prompt for the LLM can be copyright (again…human authorship), but the direct LLM output has no protection.
Yes. The rules vary worldwide. Do your own research. I did, before asserting copyright.
While it's really cool that this is possible now, this project seems to be a thin wrapper around running Claude. Presumably, someone who wanted to read one of these books could just run it through Claude themselves with comparable results.<p>Sharing some artifact of your own added value, or things you learned in the process, would be more interesting to me than the output.
Try it and you’ll find it’s about 30 hours work inside Claude to get coherent side by side markdown files starting from e.g. a single 19th c German philosophical work. At that point you can begin to edit and query the translation.<p>The problems with AI translations are a separate problem. For theoretical works they will always be somewhat irrational without a fixed glossary - which, though it sounds simple, is already an immense domain specific (really author-specific) intellectual labor involving excruciating choices for the human orchestrator - and a store of more or less trusted existing pairs (original,translation) belonging to the domain.
I can assure you this is not a thin wrapper. Even the public repo of the first workable version of it is not a thin wrapper.<p>You are welcome to try. It is not rocket science, but what I developed would take much effort to reproduce.<p>I wrote an origin story about how I got started: <a href="https://aicentral.substack.com/p/pipeline-to-the-past" rel="nofollow">https://aicentral.substack.com/p/pipeline-to-the-past</a><p>It describes some of the added value and lessons learned.
I think there's more to it than that. It's a combination of extracting the text from the PDFs, resolving persistently tricky issues (especially numbers), and flagging issues for human review at scale. You can view lessons from the author on the GitHub repo: <a href="https://github.com/sweisman/translation-pipeline" rel="nofollow">https://github.com/sweisman/translation-pipeline</a>
Why was the comment of the author of the translation-site that explains the methodology flagged / dead?<p>The mere fact that the translation was AI-assisted shouldn't be a reason to flag a comment. AI translations, especially supervised translations, can actually achieve a reasonable level of quality.<p>Disagreement on AI philosophy shouldn't be a reason to flag a contribution.
"Disagreement on AI philosophy shouldn't be a reason to flag a contribution."<p>Amen!
A professional book editor and Hebrew translator said the translation of the one work he checked out was "impressive".<p>Others have responded equally positively to the Latin translations.<p>AI as a panacea is overwrought, but there are genuine positive uses of it. Would critics rather have NO translations of these pivotal works?
Nice project, but I think the PDF format could be improved. I'd prefer having landscape layout and original text and translated text side-by-side or keep the portrait layout and do a line-by-line translation so that it is easier to review and spot translation errors.
I built Ars Astronomica to make historically important works on astronomy, calendrics, and related sciences accessible in modern English. The project combines OCR, AI-assisted translation, editorial review, and publication into a workflow that can handle book-length works while preserving diagrams, mathematical notation, and references.<p>The site currently includes English translations of rare Hebrew and Latin works, including texts that have never previously appeared in English and others that survive only in manuscript form. All editions are released under a Creative Commons license.<p>I'm interested in feedback from historians of science, classicists, medievalists, digital-humanities researchers, and anyone interested in AI-assisted scholarly publishing.
This is a valuable initiative. It would be great to include more historical context alongside each translated work.
[flagged]