An entirely Backdrop/D7-trained AI coding model: What do you think?

Hello there, i hope to bring up a time-relevant question and just leave it among the excellent and knowledgable Backdrop/D7 personnel/members who uphold the legacy and advancement of this great CMS. I am hooked, i enjoy it, and i am rooting for it. 

First and foremost, i am definitely not really an advocate for AI, since I believe that handmade things are the ultimate journey, both from gaining knowledge and the joys of progressing and getting things to work. Let alone needing to know the nuts and bolts of a project like this. The same applies to art and writing. I am more concerned about finding the golden mean these days to get things actually done.

Just recently, i have been using a certain AI code generator that would generate fully functional HTML pages (and even games) just by worded prompts. On top of that i also tried to make it generate entire Backdrop modules, which actually remotely worked (it still takes hours to explain fixes over and over again). I also prompted it to both write the code and place a few buttons to download each .info, .module file etc. so they were just needed to be placed into an FTP folder. They could be enabled. It doesn't integrate aspects such as views integrations or database interplay very well, but it was almost there.

So, that's why i was thinking about how it would be if there was someone who could train an entire AI model from scratch on the entire D7 and/or Backdrop documentation, including the latest (and other) PHP versions, SQL, JS, and all the rest of it, even D7 forum solutions, GitHub material etc. Then, let's say, it would be an universal assistant, called among the lines of "SpeedDrop" or "DropGate" (you name it), which would give accurate support in all areas around Backdrop, including the creation of modules and fixing bugs.

Since its ceiling of knowledge around the nuts and bolts of this would be set pretty high, i believe it could have potential as serving as an universal assistant for anyone who needs quicker fixes or realize unique ideas, and doesn't have too much time or energy to realize websites by hand. I personally have other endeavors calling. That's why i was wondering about how a classic project could be hybridized in a way that benefits both Backdrop itself, as well as those who use it, trust it, and expect the best of it. By hybridized i mean in a way that there is a separate assistant, while it is still up to the developer to implement the assistant's output.

Whether making such model available would require certain fees is not really important yet, but i think it even this could be recycled into the maintenance and advancement of Backdrop, too. I believe the current push and pull within the forum here is quite sufficient to get around even difficult problems, but i guess i'm just reflecting off of the worries i have around a potential case where there would come a phase that could include shortage of needed modules, or shortage of knowledgeable developers who can give advice in time. AI could bridge such gap.

So, what do you think? Hot or not?