GitHub's 2-Star AI Blacklist Gets 190 Upvotes on Hacker News
Here's what happens when someone gets fed up enough to actually do something about AI content spam. A developer named alvi-se created a simple uBlock Origin filter list on May 31, 2025, and despite having just 2 stars, 1 watcher, and 0 forks, it somehow clawed its way to 190 points on Hacker News.
The premise is brutally simple: block websites that are "filled up with ads" and pump out "mediocre content" detected by tools like GPTZero. Users subscribe by pasting https://raw.githubusercontent.com/alvi-se/ai-ublock-blacklist/master/list.txt into uBlock Origin's dashboard.
What makes this tiny repo interesting isn't its technical sophistication—it's the cultural moment it captures.
The Real Story
This isn't the first AI blocklist rodeo. We've got:
- Iz-zzzzz/Block-AI-FilterList-for-uBlockOrigin (~950 sites with an "anticapitalist focus")
- laylavish/uBlockOrigin-HUGE-AI-Blocklist (1000+ manually curated sites)
- Stevos-GenAI-Blocklist (blocks Google's AI summaries)
But alvi-se's list positions itself as the "nice alternative" to broader anti-AI crusades. Instead of blocking ChatGPT or legitimate AI tools, it targets the content farms monetizing garbage through SEO manipulation.
<> Critics argue it unfairly blocks AI-assisted content, alienating Grammarly users or non-native speakers. Maintainer's "Cry about it" stance draws backlash for dismissiveness./>
That "cry about it" attitude? Chef's kiss. Nothing says "I'm done with your complaints" quite like dismissing criticism while welcoming pull requests. It's the open source equivalent of hanging up and leaving your phone off the hook.
Detection Roulette
Here's where things get messy. The maintainer uses AI detectors like GPTZero to identify content farms. Anyone who's used these tools knows they're about as reliable as a weather forecast—sometimes right, often spectacularly wrong.
False positives are inevitable. What happens when a legitimate site gets flagged because someone used Grammarly? Or when a non-native English speaker's writing gets misclassified as AI-generated?
The Hacker News crowd is split. User lifthrasiir appreciates the categorization for network admins. Others debate whether this is a "skill issue" with AI writing tools. The consensus seems to be that while the project might be tangential to helpful AI use, at least someone's trying to solve the content farm problem.
The Bigger Picture
What's fascinating is how this 2-star repository represents a grassroots response to AI spam proliferation. No VC funding. No enterprise sales team. Just one developer who got tired of wading through algorithmic sludge.
The market implications are interesting:
1. Content farms lose traffic and ad revenue when blocked at browser level
2. Legitimate AI tools benefit by distinction from garbage
3. Ad-blockers like uBlock Origin gain relevance amid AI spam explosion
But here's the cynical reality: this is a whack-a-mole game. Content farms will adapt. They'll get better at mimicking human writing. They'll game the detection algorithms.
The Uncomfortable Truth
The real controversy isn't about blocking AI content—it's about who gets to decide what's valuable. One person's "content farm" is another's "AI-assisted productivity tool."
Alvi-se's dismissive attitude toward criticism might be grating, but it's also refreshingly honest. Most open source maintainers burn out trying to please everyone. Sometimes "cry about it" is the only sustainable response.
Will this tiny blocklist change the internet? Probably not. But it's a perfect snapshot of where we are in 2025: exhausted by AI spam, suspicious of algorithmic content, and ready to take matters into our own hands.
Even if those hands only manage to get 2 stars on GitHub.
