Settings should just facilitate getting the configuration from your dotenv file.
Rule of thumb: Each class should preferably serve a single purpose.
You should have a Database class that you can initialize with the relevant configuration from you Settings object and will handle the DB connectivity and session.
You should take a look at SQLModel and its examples.
It seems overconfidence in the output, laziness, underconfidence in writing things themselves, and offloading responsibility of thought are a few of the many reasons that come to mind. There is a minority of people who achieve better results and after gaining expertise in prompting using AI and give false hopes to the rest. Then there are the loud tech executives who promote AI with the help of amplifying media.
There is also the fact that you can quickly generate images and videos which a normal person without the proper tools and photoshop/photography knowledge would not be able to easily achieve.
I have had to spend time re-engineering AI output code. I know I’m not an expert but the output and time it takes to write long, prompts with sufficient context and detail is not trivial. A single prompt is not sufficient. Better outputs are achieved through several structured prompts for every “job” in a “team” needed to achieve the outcome (product owner, project manager, software engineer, quality engineer, UI engineer, etc). for writing articles or a book, I would expect a similar set pattern to achieve better results.
And then there is the whole thing about “hallucinations” which undermines quality.
Statement vs question