It includes a lot of steps and constant human evaluation between them, which implies that decisions about tone are ultimately made by whoever is prompting the LLM, not the LLMs themselves.
> "If they are generally in the style I want..."
> "choosing the sentences and paragraphs I like most from each..."
> "I also make my own adjustments to the translation as I see fit..."
> "I don’t adopt most of the LLM’s suggestions..."
> "I check it paragraph by paragraph..."
It seems like a great workflow to speed up the work of an already experienced translator, but far from being usable by a layman due to the several steps requiring specialized human supervision.
Consider the scenario presented by the blog post regarding bluntness/politeness and cultural sensitivities. Would anyone be able to use this workflow without knowing that beforehand? If you think about it, it could make the tone even worse.
Just like programming. And anything else "AI" assisted.
Helps experts type less.
I worked with Java for a couple of years. Lots of boilerplate required. The IDE handled them just fine. I never felt I needed something smarter in order to write verbose stuff.
My style is completely different though. Whenever I can, I will choose languages and technologies that don't require boilerplate.
https://wiki.c2.com/?BoilerPlateCode
However, this is deviating from the subject (look at the original post again). We're not talking about dumb, automated boilerplate code. The discussion is much deeper, drawing parallels with tone in translations. We're obviously talking about non-boilerplate.
Can you give us an example of a typical translation question and the "good prompting" required to make the LLM consider tone?