I don't particularly care if vibe coding and the like are used for web apps and mobile apps. The quality there has always been poor and gets worse over time. AI slopware is just the new low and in a few years time I'm sure they will find a way to make things even worse.
But for software infrastructure; Kernels, operating systems, compilers, browsers, etc, it is crazy we are even considering AI at it's current ability. If we are going to do that, we need to switch to Ada/SPARK or some other type of formally verifiable system.
Maybe I'm overreacting, but all I want to do right now is escape. It horrifies me to think that one day I may be driving a car with a braking system vibe coded in C++.
Great care and attention is required for critical system components and LLMs lack both.
Not to mention the copyright risks - do we really want a piece of code that can’t be licensed or turns out to be a verbatim copy from another project to end up in the kernel or something? (No, the answer is we don’t want.).
i dont work in then auto industry but Ive read stuff from people that do and Im pretty sure I remember they all say that all major car mfg code is tons of auto generated slop even pre AI.
I've worked as a software engineer with different types of engineers (electrical, mechanical and automation).
Their testing is often more strict but that is a natural consequence of their products being significantly harder to fix in the field than a software product is.
Other than that, my experience is that our way of working on projects across disciplines is very similar.
There are many camps. Some programmers embrace the prompt, some use parts of it, some reject it on principle (a dying breed?), some think that non-developers are finally getting a go at it, some gloat that tech barons are making software engineers (apply optional quotes) obsolete.
It’s all too varied to put people into one or two camps.
How big a problem is that over a multi-decade time horizon?
There is a pretty big variation in average temperatures by country [0]. Somehow People everywhere from Thailand to Greenland manage to find food. All else failing it is a possibility to trade for calories. Let alone technology improvements that might save the day by accident.
I mean, it might make places uninhabitable over the course of a few generations, but things that change so slowly are't actually much of a threat on an individual level. Worst of the worst cases people can move or not have children - the statistics suggest that is an acceptable option to a lot of people.
Right, and we know that modern human societies are really good at planning for a major disaster on a multi-decade time horizon! Look at how well we've dealt with the climate cri—oh, wait.
This won't end humanity, no. But it is likely to cause absolutely catastrophic levels of upheaval and probably billions (with a b) of deaths—from famine, disease, exposure, and war.
> Multi-decade time frame can only be thought of as catastrophic when it comes to change of this magnitude.
If someone moves from Singapore to Poland is that a catastrophe? We'd be talking a smaller temperature delta than that and this is still a theoretical risk. That isn't necessarily something that people worry about beyond saying "it is very hot" or "it is very cold". It doesn't have a lot of implications beyond needing to move crops around and changing building standards (which is achievable over long periods of time).
Now just a 10-15 degree swing isn't the end of the story because if the average temperature crosses 0 that might well be a big problem. I dunno. But it sounds solvable, these aren't particularly scary scenarios being put forward. They're more of the expensive and inconvenient variety.
Livestock, staple crops, and pollinating insects cope pretty well over a wide variety of temperatures. Some specific crops don't, but that's not a problem as long as changes are predicted.
I only just learned about SunCable. I think using our vast swathes of empty, sun-drenched land to provide power to our Southeast Asian allies is a great idea.
Programmers tend to lean two ways: math-oriented or literature-oriented. The math types tend to become FAANG engineers. The literature oriented ones tend to start startups and become product managers and indie game devs and Laravel artisans.
Lots of workshops, factories, university research labs, etc. still use old machinery that would be a huge waste of money to replace just because the computer that controls it runs Windows 95. In some cases it can't be replaced because the company that created the software, drivers, or IO cards is long gone.
PUC Lua is supposedly a bit of a pain for ffi, but I havent tried it myself. Luajit is some kind of crazy magic. You can (almost) just copy and paste the c header file into the ffi.cdef function and then start using c functions as if they were lua functions.
Maybe you don't read much, but it's obvious they weren't making some universal statement about code. They are referring to the code you write when you are just experimenting by yourself, for yourself. The point is to not let irrelevant things like usefulness, quality, conventions, etc. limit just tinkering and learning.
But for software infrastructure; Kernels, operating systems, compilers, browsers, etc, it is crazy we are even considering AI at it's current ability. If we are going to do that, we need to switch to Ada/SPARK or some other type of formally verifiable system.
Maybe I'm overreacting, but all I want to do right now is escape. It horrifies me to think that one day I may be driving a car with a braking system vibe coded in C++.
reply