None of the "code generation" stuff is new by the way.
-
None of the "code generation" stuff is new by the way.
The tech industry has tried to speed up coding and increase software output for the last 3 to 4 decades, by various means; e.g. Rapid Application Development, Expert Systems, Object-Oriented Programming, thousands of different frameworks all the way to trying to off-shore development and exploit third-world labor.
The problem with this is: there is no software scarcity. Pretending that "we can't make software fast enough" is a red herring to hide the fact that making (good) software is 90% painstaking research, design, planning, marketing and talking to and supporting customers.
And 10% writing the actual code—the C-suite is doing ye olde "trying to find a technical solution to a social problem".
Exactly. What bothers me isn't code generation. That's a good idea if done correctly (with precise tools).
What bothers me is the technofascist makeover of our world.
-
None of the "code generation" stuff is new by the way.
The tech industry has tried to speed up coding and increase software output for the last 3 to 4 decades, by various means; e.g. Rapid Application Development, Expert Systems, Object-Oriented Programming, thousands of different frameworks all the way to trying to off-shore development and exploit third-world labor.
The problem with this is: there is no software scarcity. Pretending that "we can't make software fast enough" is a red herring to hide the fact that making (good) software is 90% painstaking research, design, planning, marketing and talking to and supporting customers.
And 10% writing the actual code—the C-suite is doing ye olde "trying to find a technical solution to a social problem".
@thomasfuchs I don't think this is the entire story. Tools and techniques like RAD/OOP/Expert Systems/4GL can definitely save time when used correctly. Abstracting or automating boring parts leaves more time and headspace for the complicated parts -- which are typically the business rules and the non-functionals.
The way LLMs generate code is the exact opposite: they make it harder to focus on the hard parts by trying to generate "everything".
-
@thomasfuchs I don't think this is the entire story. Tools and techniques like RAD/OOP/Expert Systems/4GL can definitely save time when used correctly. Abstracting or automating boring parts leaves more time and headspace for the complicated parts -- which are typically the business rules and the non-functionals.
The way LLMs generate code is the exact opposite: they make it harder to focus on the hard parts by trying to generate "everything".
@elricofmelnibone maybe these tools save time or improve quality, maybe they don't, it probably depends on circumstances.
but my point is: it doesn't matter if you can speed up 10% of the total effort to make software by 5%; that's a rounding error.
-
@elricofmelnibone maybe these tools save time or improve quality, maybe they don't, it probably depends on circumstances.
but my point is: it doesn't matter if you can speed up 10% of the total effort to make software by 5%; that's a rounding error.
@elricofmelnibone what actually happens is that the important parts of software development are starved of attention because "we can write software so easily now"
-
@alper
@thomasfuchs it's also where they're as likely as not to take what should be a footgun and turn it into a self-inflicted head shot. -
@dymaxion @thomasfuchs It’s been debugging DNS and other issues for me just fine. Ideological and outdated views here aren’t going to be very productive.
-
None of the "code generation" stuff is new by the way.
The tech industry has tried to speed up coding and increase software output for the last 3 to 4 decades, by various means; e.g. Rapid Application Development, Expert Systems, Object-Oriented Programming, thousands of different frameworks all the way to trying to off-shore development and exploit third-world labor.
The problem with this is: there is no software scarcity. Pretending that "we can't make software fast enough" is a red herring to hide the fact that making (good) software is 90% painstaking research, design, planning, marketing and talking to and supporting customers.
And 10% writing the actual code—the C-suite is doing ye olde "trying to find a technical solution to a social problem".
@thomasfuchs yup, we used custom Fortran pre-compilers to build complex numerical simulations of nuclear power plants way back in the 1980s. They were error-prone and had to be manually debugged, but it was still considered a major advance over attempting to do all that programming by hand.
-
M monkee@other.li shared this topic
