How does it compare to previous (human) translations, what is the regurgitation level? In any case, lots of em dashes, so mission accomplished to bring it up to contemporary standards.
Emily Wilson's translation is way better, more modern and adequate to present-day readers and, probably more context aware.
<a href="https://theclaudyssey.com/read/book-01.html#n1" rel="nofollow">https://theclaudyssey.com/read/book-01.html#n1</a><p>The first line is a crappy mishmash of Lattimore and Fagles.<p>Lattimore:
>Tell me, Muse, of the man of many ways, who was driven far journeys, after he had sacked Troy’s secret citadel.<p>Fagles:
>Sing to me of the man, Muse, the man of twists and turns
driven time and again off course, once he had plundered
the hallowed heights of Troy.<p>Claude:
>Tell me the man, Muse — the man of many turnings, who was driven wandering far, once he had sacked Troy's holy citadel<p>Combining "man of many ways" and "man of twists and turns" into "man of many turnings" is just dreck.
The sort of sad but interesting thing here is that while this claims to be a more direct translation, Claude models were certainly trained on all of the pre-existing translations and this is likely going to directly influence the output in the ways you've outlined here. I'm not sure any modern LLM could honestly be claimed to produce a translation that doesn't draw on these pre-existing sources.
The only thing Fable was given was a clean of copy of the ancient Greek. This is not a bastardization of other translations. I appreciate and value the critical feedback in any case (truly).<p>I find that Fagles is very majestic -wonderful, really- but he's also very liberal (poetically) and definitely did not limit himself to the same number of lines as Homer as this project did.
As another comment here points out, Fable likely had multiple translations, academic literature about the text, and maybe the original as well, in its training set. So even if only the original Greek was given to it as input it probably leveraged prior art when translating it.
> The only thing Fable was given was a clean of copy of the ancient Greek<p>Ok, but this clearly combines a few translations line by line if you work through it.
You think Anthropic didn't train on some of the most famous translations of the Odyssey? Really?
This is a useful marker of progress in translation, not a replacement for literary translators.<p>Six years ago I fed Dante's Inferno to AWS translation. It was...not good[1]. Maybe its time to try again and see how things have changed.<p>[1] <a href="https://github.com/ffx271/dante-inferno/blob/master/english/txt/aws/en.canto01.txt" rel="nofollow">https://github.com/ffx271/dante-inferno/blob/master/english/...</a>
12,107 line-for-line translation of the A.T. Murray Greek text (Loeb, 1919) by Claude Fable. Includes line notes and an index. HTML, epub, pdf, and YouTube Audio (by gpt4o-TTS). A dual-agent review pass was done on the entire work with a third "adjudication" agent reconciling their findings. Reviewer agnets were gpt-5.6-sol high reasoning.
why?
One very simple benefit is that it's unencumbered by copyright.<p>I do value the work of human translators; when they're really in harmony with the original work, they can be fantastic. But they can also be disastrous, like Alfred Birnbaum's translations of Murakami, which radically reshapes the protagonist and threw out nearly 20% of the original work.
If the Murakami was undertranslated then how will having more translations that crowd out the potential revenue of future translators help produce better quality translations?
The problem is that (we have every reason to believe) models like Claude were <i>trained</i> on existing translations. Now they're being used to generate things that are direct replacements for those translations in the marketplace.<p>This is basically an anti-AI author or lawyer's dream case.<p>And, seeing this uncanny-valley-of-slop output, I kind of hope they go for it.
Every translator is opinionated in some way. I can see the appeal in having something like this as a neutral comparison.
LLMs are not neutral, either. They don't have "opinions", per se, as they don't want anything, but they do have biases baked into their training.
I love the poem and wanted a modern translation with annotations and index, etc., to be freely available to anyone who wanted to read it. I had a sense that Fable would do a really good job. I am pleased with how it came out. The review agents didn't change a single word of the translation, only some of the notes.
This could allow a very opinionated and tailored translation, which I think is the appeal.<p>That said, I'd like to see the prompt(s) having such an effect on the subjective choices for a project like this.
This page provides a great overview of several popular translations of The Odyssey. It lists line-by-line translations as well free translations:
<a href="https://readtheodyssey.org/best-odyssey-translation" rel="nofollow">https://readtheodyssey.org/best-odyssey-translation</a>
Read this as "Claudussy" at first.
Having recently read Wilson's translation, I was immediately put off at the introduction and leapt to the conclusion it was bad. It just doesn't sound anywhere near as good to my ear.<p>However, there may be more to it than I thought. No admittedly the following is ChatGPT's opinion, but I think it's valid and made me think again:<p>> The second is the better translation as English poetry; the first (Claude's) is the better translation if you want to see the Greek more directly.<p>> If I were studying Homer, I'd strongly prefer the first. It continually reminds you that you're reading an ancient poem whose concepts don't map neatly onto modern English.
Translation is an art form that is rapidly becoming lost, not because AI is doing it, but because there are too many translators who have apparently been trained (or likely, were never trained in it as a formal study at all) that it should be written in a way that emphasizes the translation as a standalone work more than anything else.<p>A good translation should attempt to communicate what the author meant, <i>as closely as possible</i>, even if there is no equivalent in the target language.<p>A good translation <i>should</i> be beautiful all on its own.<p>It's the balance of the 2 that makes a translation successful.<p>If you erase the author's true meaning just because you don't want to slow the reader down in having to take a moment to read an explanatory footnote if they're confused, you have failed to translate it, and have likely editorialized it.<p>Why do many English-speaking people know 'wabisabi' as a concept? Because someone bothered to explain it to them, and they bothered to listen. That act enriches the reader.<p>And don't even get me <i>started</i> on "localization".
Probably this will be done more and more often for works which don't have translation in some language because of low interest. E.g. old manuscripts, light novels, fan fiction etc.
I don't remember so many emdashes in the original.
I'm just curious why you'd put "Claud[e]" in the name of this project. Is it an Anthropic-sponsored one? If so, may you please disclose it? Additionally, I think it puts a date to a wonderful project that otherwise could be very timeless. In 5 years we'll probably be like, "what's Claude?" Anyway, really cool work and thanks for sharing!
I truly doubt it will be timeless. There are already a few excellent English translations of the Odyssey.
It is just a domain for the project's github page. All the files, YouTube uploads, etc., are The Odyssey or Homer's Odyssey.
In 5 years we'll probably have locally run AI which could do something like this in a millisecond.
A language model does language model things? That might actually be... an appropriate use of the technology!
What is the point of this?
Now have it make it into a movie.