aside from the obvious use of LLMs for the actuall project, everything was written by hand … everything you see is my fault.

deleted by creator

assert(true);

Good point

I do, but something about it doesn’t feel right. Maybe I’m just not committing to it cos I can too easily jump back to the host OS

Usually if I’ve made a mess with installs or config and want a clean slate, but this time I’m doing it cos I tried out a new OS and it didn’t work out.

Hard to say exactly, how much, maybe twice a year, but when I do reinstall I often reinstall like three times in quick succession until I get a good foundation.

It’s not only LLMs that are problematic. Vision models, for example, present severe mass surveillance problems. Nor are tech bros only involved in LLMs.

Thanks but not considering hardware adjustments. I’ll look at the bios board efficiency, turning it off, and the wall meter thing

desktop … lots of browsing, lots of coding/compiling, some gaming

Here’s the list

• AMD 9950X
• Deepcool AK620
• MSI B850M GAMING PLUS
• TEAMGROUP 2x32GB DDR5-6000 CL38 Memory
• KIOXIA EXCERIA PLUS G4 2TB PCIe 5.0 SSD
• AMD 9060XT 16GB GPU
• Jonsbo D32 PRO MESH MicroATX Case
• Corsair SF850 PSU
• 3x ARCTIC P12 120mm Fan
• 1x ARCTIC P14 140mm Fan

part list here https://programming.dev/post/52768664/24715637

I suppose I mean “which is most efficient out of the box?”

Elon Musk

thanks for the great technical suggestions everyone

well I do hate going into town, so that would work

reddit; may add news sites given the wonderful news we’re getting these days