zlacker

llms are already very expensive to run on a per query basis. Now it’s being asked to run on massive codebases and attempt to fix issues.

Spending massive amounts of:

- energy to process these queries

- wasting time of mid-level and senior engineers to vibe code with copilot to ensure train and get it right

We are facing a climate change crisis and we continue to burn energy at useless initiatives so executives at big corporation can announce in quarterly shareholder meetings: "wE uSe Ai, wE aRe tHe FuTuRe, lAbOr fOrCe rEdUceD"