Overview
What The Internet Loves is a discovery almanac. It shows what people are searching, building, watching and discussing right now, gathered from public sources, with every figure labelled by where it came from.
The aim is to make the current internet easier to browse without pretending that a number on one platform means the same thing on another.
The problem
Each platform shows its own feed. Working out what is rising across search, code, video and discussion means opening several sites and comparing numbers that were never measured the same way. It is also easy to mistake two things happening at once for one causing the other.
How it works
- Live sources, read on the server. Hacker News, GitHub, Google Trends' public daily feed and Bluesky's trending topics are loaded through their public interfaces. YouTube, Reddit and Product Hunt connect once their access credentials are configured.
- Honest source states. A source that is not connected is simply not shown. Nothing is filled in to cover a gap.
- Summaries you can check. Items and tabs can be summarised by an open-source language model. The model is given only the items on screen, and each summary links to its sources.
Principles
- Origin on every figure. Each card and chart says whether its data is from an official API, a public feed, or sample data.
- No cross-platform arithmetic. A GitHub star, a YouTube view and a Reddit upvote are never added together or ranked as if they were the same.
- Correlation is not cause. Topics that appear on several platforms are described as appearing together.
Built with
Next.js, TypeScript, Tailwind CSS, Recharts, Framer Motion and Lucide icons, with server-side data adapters and an OpenAI-compatible model client.