ZeroHour
DashboardNewsCVEsExploitsIncidentsAIBriefAnalyticsEntitiesWatchlistIOCsAsk the feedSourcesFeeds & APIKiosk mode
Product

Hacker News

2 mentions in 7 days · 2 in 30 days · 2 total · first seen · last

Search everything

Timeline

How Hacker News ranking works: scoring, controversy, and penalties (2013)

A 2013 analysis of Hacker News reveals its gravity-based ranking formula and hidden penalties on NSA and controversial stories.

A detailed study of the top 60 Hacker News stories over several days in November 2013 found the published ranking formula is mostly accurate but supplemented by extensive manual and automatic penalties. Roughly 20% of front-page stories were penalized, including an automatic 0.4 penalty for any title containing "NSA" and severe "controversy" penalties after 40 comments. Popular domains like arstechnica.com, medium.com, and github.com received 0.25-0.8 domain penalties, and stories are reranked only when upvoted or via a random 30-second cycle among the top 50.

Hacker News · security · 5h agoOtherHN 21↑ · 2 comments

How come AI-related posts get so many points on HN?

Hacker News users debate why AI posts dominate the site, attributing engagement to hype cycles, mixed enthusiasm, and possible vote boosting rather than quality.

A Hacker News Ask thread with 18 comments discusses why AI-related posts consistently attract points and engagement. Commenters compare the current AI wave to earlier crypto and Twitter hype cycles, with enthusiasts, critics, and fatigued readers all joining threads. Some users speculate about possible astroturfing or boost campaigns by AI companies, and note that AI projects receive outsized GitHub stars while research papers from DeepSeek and DeepMind draw less substantive discussion.

Hacker News · security · 20h agoOtherHN 20↑ · 18 comments

Appears with

ai model DeepMind 1ai model DeepSeek 1organization Hacker News 1organization OpenAI 1organization Y Combinator 1vendor Hacker News 1vendor Hugging Face 1vendor Y Combinator 1
Entities are extracted by the model from each article. Watching an entity keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.