99% of code scriptmonkeys write for web and games could be easily written by simple neural network trained on their code.
99% of code on github is scriptmonkey crap. So you could easily deduce what their neural network will produce trained using code from github.
Of course, none of that code should not being wrintten in the first place, but then hardware corporations would had no chance to sell their new hardware. So, bloatware is necessary for their survival, but they don't want to pay scriptmonkey salaries for it.
And as always with anything the "garbage in, garbage out" principle works like a charm.
99% of code scriptmonkeys write for web and games could be easily written by simple neural network trained on their code.
99% of code on github is scriptmonkey crap. So you could easily deduce what their neural network will produce trained using code from github.
Of course, none of that code should not being wrintten in the first place, but then hardware corporations would had no chance to sell their new hardware. So, bloatware is necessary for their survival, but they don't want to pay scriptmonkey salaries for it.
And as always with anything the "garbage in, garbage out" principle works like a charm.
Really, I'm amased they didn't do that years ago.
Writing code is only 50% of the job.
Gathering requirements is the difficult part.