jonny (good kind)People already are using "i saw this video before 2023" as a signature that the video is eligible to be believed. Soon, "this software was developed before 2025" will be the only sign that makes software eligible to be run.
when a project is developed in such a way that every line of it might be changed tomorrow in incomprehensible ways, how could you ever trust that software?
this is not a hypothetical, this is how i evaluate code right now. if it is perfect boilerplate with sole contributor and repo creation date >mid 2025, i do not run that code
The Doctor@jonny So those of us shouldn't bother starting new projects? Because we'll spend all our time defending our status as organics rather than hacking code?
This makes me wonder if it's even worth working on OSS anymore.
@drwho
I'm realizing my off the cuff thoughts here are being taken a bit more literally than I intended - there's a whole aura to AI code that isn't just timestamp and contributor count, its pretty easy to tell when something was written by a human being. I was mostly commenting on the similarity between what's happened to perception of video and perception of code - that in general post AI things have a higher threshold of trust and require closer scrutiny. I certainly would never lead nor encourage some witch hunt to prove humanity or vice versa, just like if I find some random GH repo lying around I take a closer look at who did it and how it was written.
tl;dr not giving up on FOSS