PSA Huggingface Is Scraping GitHub

HuggingFace out here playing 4D chess with developer data. They're scraping every GitHub repo for LLM training datasets, which is already questionable enough. But the real galaxy brain move? Making it opt-out instead of opt-in, then staying suspiciously quiet about it. So if you want your code excluded, you have to manually open a GitHub issue like some kind of peasant filing a support ticket.

But wait, there's more! Even if you request removal, they're storing everything in Git history. You know, that immutable ledger we all love that makes deleting things basically impossible? Yeah, those "redaction" commits don't actually purge your data—it's all still there in the commit history, forever immortalized in the blockchain of shame. So your code is getting fed to AI models whether you like it or not, and the removal process is about as effective as using git push --force to erase your embarrassing commits from your coworker's local clone.

The escalating Vince McMahon reaction perfectly captures the journey from "meh, data scraping" to "WAIT WHAT" as each layer of this dystopian onion gets peeled back.