Tool tracks AI mentions in Hacker News story titles using Algolia and Firebase APIs

Tracking AI Mentions on Hacker News

A new tool has been created to quantify the appearance of AI-related terms in the titles of Hacker News story submissions. This utility monitors new posts from midnight UTC onwards, providing a real-time count of how often artificial intelligence is referenced.

Methodology for Term Identification

The tool employs two primary methods for identifying AI mentions: a 'strict' count that looks for the standalone word 'AI' (case-sensitive and word-bounded, excluding terms like 'OpenAI'), and an 'extended' count that includes 'artificial intelligence' and a list of specific model and vendor names (case-insensitive). This dual approach allows for a nuanced understanding of AI topic prevalence.

Data Sourcing and Real-time Updates

Initial data for today's figures are sourced from Algolia HN Search. Subsequently, the tool stays current by streaming new stories from the HN Firebase API, ensuring that counts are updated within seconds of a story being posted. A reconciliation check runs every five minutes to prevent data discrepancies if the stream encounters issues, with a fallback to polling every 45 seconds if the stream fails to open.

Historical Data and Privacy

The tool also provides historical data, with a range chart counting title matches per UTC day and month back to 2007, prebuilt at deploy time. The application runs entirely in the user's browser, with no server or database, and no user data is stored or cookies set. Page views are counted by Cloudflare Web Analytics, which is cookieless.