5 ms·
> That map makes it stupidly easy to tell a human from a machine. Someone who commits 50,000 lines in an afternoon did not type those by hand. Someone doing a c
by WCSTombs 1mo ago
> That map makes it stupidly easy to tell a human from a machine. Someone who commits 50,000 lines in an afternoon did not type those by hand. Someone doing a couple hundred? Probably real. Simple.
Uh, this isn't true, though? I squash commits all the time locally before I publish them. There can be many reasons for that, but at the end of the day, I'm not obligated to share my messy revision history with the world. In the most extreme case, the first commit on one of my projects spans multiple years of development because it was forked from another project that's now pretty irrelevant to it. On the other hand, I've received pull requests that were a couple dozen lines and were definitely LLM-authored.
As sort of a counterpoint to the author's goals, IMO the community should extend the benefit of the doubt to writers and artists who explicitly say they don't use AI, because it's just really hard to prove that you're not using AI. Trying to force any kind of system will inevitably affect the quality of what you make, and the art community should not want that.
- bradyd 1mo agoIt's also possible to break the 50,000 lines up into multiple commits and fake the commit dates, so this does nothing to prove a human wrote them.
- Krutonium 1mo agoYea there's literally software that lets you put art into your commit history blog graph on github by making a repo with potentially tens of thousands of dummy commits.
- KolibriFly 1mo agoSquashing commits pretty much turns the whole "green carpet" into meaningless noise