On 12/12/25 16:38, Dave McGuire via vcf-midatlantic wrote:
I've not used it in any of my professional work, but I've been experimenting recently with what the kids call "vibe coding" for generating boilerplate code and test functions.
The vibe coding scares me, I don't think I understand it. It is just not sinking in. Might be the tester in me. :-) I just tried the same asm 68K code query that triggered the initial post with Claude, code is better but Gemini made the assumption that the DUART had a 3.6864 MHz Crystal. Claude doesn't state this assumption but does get matching settings. I didn't state that in my initial (very vague) requirements. So, +1 for Claude. :-)
It's actually surprisingly capable. You treat it (and talk to it) like it's a very green intern, and you have to check and often correct everything, but I've seen how it can be a timesaver. Sometimes a significant one. Capable but it has limits. It is just that AI isn't living up to the marketing hype in my eyes and I'm trying to figure out if I'm doing something wrong. The Linux Flex disk programs were a terrible mess. Took me a week to figure out how it wasn't doing the math.
I don't like to feed suits' greed with my personal info, so I've not used any of the online AI "services". I built a system with a 16GB GPU (Nvidia P100) in an HP DL380 server, and I run Ollama and OpenWebUI on it. Overall it was pretty easy to put together, and while it's surely nowhere near as capable as the online services with tons of memory and giant models, again I won't feed suits' greed with yet more of my personal information.
Actually that is what a lot of us are doing. We're 'training' the AIs for the big corporations. Scares me. I'd love to setup a home AI. Just have too many other things to do.
won't feed suits' greed with yet more of my personal information.
+1
The key is to always remember that half of the results will be wrong. You have to understand the code it's writing, and the language in which it's writing it. The business world has a massive collective hard-on for using it to get rid of all those troublesome expensive tech people with their lack of ties, lack of golf, long hair, and crazy ideas about being the real value in a corporation, but that won't happen en masse anytime soon without huge, corporation-killing messes...some of which have already started. You MUST have the discipline to not just take what it writes verbatim. You MUST review it, every single line, and treat it as what it actually is: a starting point.
I don't trust it, that's why I have tests.
But within those limitations, I've seen with my own eyes how even small models (< 16GB) can write a lot of boilerplate drudge code and test harnesses, to the point of actually saving real time. That's were I may need to change. Although I haven't liked a lot of the software design (functions et al). I tend to think a certain way (high level to nitty-gritty details).
But i do agree that I can use it for the boilerplate parts. I may need to revise my thinking in that direction.
As with any tool, it must be used within its limitations, which of course means understanding those limitations. I'm trying to do this but they keep upgrading the capabilities.
For my own work, it's not likely to be extremely productive soon. It was trained on mainstream trendy stuff, which is sloppy Javascript and gigantic loads of webby UI code. I primarily write firmware for laboratory instrumentation and communications systems: the sort of stuff that very few dorm room rats are barfing out. I can't tell you if my code is professional. :-)
-- Linux Home Automation Neil Cherry ncherry@linuxha.com http://www.linuxha.com/ Main site http://linuxha.blogspot.com/ My HA Blog Author of: Linux Smart Homes For Dummies KD2ZRQ