The new translation is still underway, initially from what I can tell the results look more high quality than the current version. However, the current one stands still well on its own as being understandable enough.

That said, as a intro prototype, I took a dive into the keyword of ‘Jesus’ pbuh, under the current system, in order to extract stories from Tafsir related to him. Sometimes, and not often, I do scratch my head on wondering about some of the translations. If you are a Muslim, it pops out where it might be odd. But I have come to realize:

–

1. I could be intentional. Or,
2. It could have been an OCR issue, and that is just what it looked like to the OCR, and thus, the AI isn’t wrong on what is is translating. Or,
3. The context is broken, and without a reference point sometimes in translation it can cause an odd situation. Or,
4. The AI is hallucinating, which I did catch a few times it churning out odd results. I even caught it engaged in a vicious endless loop more than once. But it seems as if the current model I’m running is handling this much better.

–

Now, furthermore, I reached out to several publishers and different groups to see if they are interested in working with such a project, but apparently it appears to be a quick denial or ignoring. Nonetheless, The prototype is a success, and I don’t see anyone with such a translation to review right now. If anyone is interested in these types of projects, my e-mail is on the main site.

–

In the meantime, I have some other projects and topics I’d like to organize and compile here, without any particular timeline, and I’ll likely do that pending the v3 translation set.

–

The last item of amusement is, for a new site, watching the bots try to login is amusing. It’s wildly impossible to do, but it goes to show most of the traffic on the internet is junk. Spam, Ads, Tracking, Bots, far exceeding meaningful purposes.