All that effort and the writing of very specific prompts in very specific ways in order to create a determenistic output just feels like a bad version of a programming language.
If we're not telling the computer exactly what do then we're leaving the LLM to make (wrong) assumptions.
If we are telling the computer exactly what to do via natural language then it is as complicated as normal programming if not more complicated.
One of the most frustrating (but common) things is you do v1. It looks good enough.
Then you go to tweak it a little (say move one box 10-15 pixels over, or change some text sizing or whatever), and it loses its mind.
So then you spend the next several days trying every possible combination of random things to get it to actually move the way you want. It ends up breaking a bunch of other things in the process.
Eventually, you get it right, and then never ever want to touch it ever again.
These are all just stopgaps, this tech is still in its infancy. If it keeps improving, it will reach a point where it can implement complex things from simple prompts, the way that talented programmers can.
I’m not a fan of a lot of this AI stuff, but there is no reason to expect it won’t get to that level.
Talented programmers often don't get things right though. They can make the wrong assumptions about what a product person wants or what a client wants.
And that normally stems from lack of information or communication problems.
It’s magical thinking to think it won’t! We already have one example of it being possible (in humans). Unless you think humans have a “soul” or some other intangible element, then there’s no reason they can’t be emulated.
If we're not telling the computer exactly what do then we're leaving the LLM to make (wrong) assumptions.
If we are telling the computer exactly what to do via natural language then it is as complicated as normal programming if not more complicated.
As least that's how I feel about it