Is this something you would run locally with Gemma 3?

I came here to say what @bnuyaminim.bsky.social already pointed to: decide whether you're asking about sound or its visual representation.

Fulfilling Jesus' reading of Isaiah

Yes, and since it's imperfect rather than aorist, I'd prefer "kept weaving".

IIRC, Young and Rezetko represent the minority view that Hebrew diachrony is impossible to disentangle. Hurvitz represents the conventional periodization. Hornkohl adds a transitional phase to Hurvitz's early & late phases.

It seems to me Aaron Hornkohl's views represent the scholarly consensus.

I think there's some misunderstanding of what Hugging Face is. HF is a repository of models of various origins. It's like the archive.org or GitHub of AI. Van Strien's actions are not the actions of Hugging Face. Here's an HF-independent source: arxiv.org/abs/2211.05100

The intellectual property issue: it's currently possible for training to use only ethical sources. See arxiv.org/html/2412.06...

Here's some evidence. This Greek text has not previously been translated into English. kmpenner.hcommons.org/2024/03/19/w...

Some differences: ancient language (no longer spoken, limited corpus); copious secondary literature, plenty of grammatical and lexical publications.
LLMs are good at translation into English; some chatbots produce translations better than graduate students, and are orders of magnitude faster.

*willing=wrestling

I appreciate your input. Besides the immediate chapter I'm writing, I'm also willing with our responsibility at this point.
Should we be researching ways to make AI ethical, or should we be opposing its adoption? Or both: oppose until it's ethical?

As for accuracy, I expect we agree that ideas don't have to be perfectly accurate to be ethical in the classroom, or we wouldn't be allowed in!
But what is our error tolerance?
Chatbots are improving; will they ever pass that threshold?

Fair. Do you mean There IS no ethical classroom AI use (as things now stand), rather than There CAN BE no ethical classroom AI use (ever)?

I'm interested because I've been asked to write a chapter evaluating the usefulness of AI for Bible translation.
What if a LLM were trained without stealing IP and without excessive energy consumption?
What if it's used for things other than replacing thinking?
What if it becomes more accurate?