Project ideas from Hacker News discussions.

What's Earth's dominant species by mass?

📝 Discussion Summary (Click to expand)

Theme 1: Biomass and population comparisons
Commenters repeatedly compare the numbers, mass, or prevalence of different life forms (bacteria, whales, pandas, humans, etc.).
- “More than 2/3rds of the world's avian biomass is poultry” – ortusdux
- “Individual whales are large, but there aren’t that many individuals.” – dhosek
- “All the humans on the planet could fit in a cubic hole 1/2 mile on each side.” – niccl

Theme 2: Humorous analogies and whimsical takes
Many responses turn those comparisons into jokes or absurd imagery (panda‑express vs. pandas, slip‑n‑slide pandas, beetle‑filled deities, etc.).
- “Slippery. Now I want to give a panda a slip‑n‑slide and film it.” – komodo99
- “Famously, when asked about a god, J.B.S. Haldane allegedly said that if there is a god or gods, they must have an ‘inordinate fondness for beetles.’” – gumby
- “How does ‘There are more jiffy’s lubes than pandas in the world’ sound?” – anukin

Theme 3: Language evolution and phrase contraction
A smaller but notable thread discusses how English phrases contract over time, speculating on the future of odd constructions like “gave human written.”
- “English words like 'goodbye' are contracted from longer expressions after centuries of common use[0]. Reading your phrase 'gave human written' made me wonder how this phrase would contract if it were still in use centuries from today.” – elevation
- “It was fun and gave human written. Thanks for that!” – gkoenig
- “funny that i didn’t think about it till i read your comment. and then realized i read the entire thing. checks out.” – apsurd


🚀 Project Ideas

AI Landing Page Detector for HN

Summary

  • Detects whether a linked landing page on HN is likely AI‑generated using linguistic heuristics and a lightweight ML model.
  • Core value: saves users time by flagging AI‑generated content, letting them skip low‑effort pages.

Details

Key Value
Target Audience HN readers who feel overwhelmed by AI‑generated landing pages
Core Feature Browser extension / userscript that adds an “AI” or “Human” badge next to each HN post link
Tech Stack JavaScript (Chrome/Firefox extension), TensorFlow.js or a small hosted model (e.g., HuggingFace DistilBERT)
Difficulty Medium
Monetization Hobby

Notes

  • “I don’t think I’m physically capable of reading the AI landing pages of every new product on HN.” – apsurd; this tool directly addresses that frustration.
  • Could spark discussion on transparency of AI‑generated content and encourage creators to label their pages.

HN TL;DR Summarizer Bot

Summary

  • Generates concise summaries of linked articles (especially landing pages) directly in HN comments via a bot or bookmarklet.
  • Core value: lets users quickly gauge relevance without opening each link, reducing fatigue.

Details

Key Value
Target Audience Busy HN readers, researchers, product managers
Core Feature Given a URL, returns a 2‑3 sentence summary using an LLM or extractive summarization; shown as a comment or tooltip
Tech Stack Backend: Python (FastAPI) with summarization model (e.g., PEGASUS/BART) or OpenAI API; Frontend: HN comment bot via HN API or userscript injecting summary box
Difficulty Medium‑High
Monetization Revenue-ready: Subscription ($5/mo for premium summaries, free tier with limited usage)

Notes

  • apsurd’s comment about being unable to read all AI landing pages highlights the need for quick consumption aids.
  • Beyond HN, the summarizer could be useful for any link‑sharing platform, encouraging broader utility and discussion about summary quality.

HN Feed Filter & Mute Manager

Summary

  • Allows users to customize their HN feed by muting or downranking posts based on keywords, domain, or AI‑generated likelihood.
  • Core value: reduces noise, lets users focus on content they care about.

Details

Key Value
Target Audience Power HN users who want control over their feed
Core Feature Settings web app where users define rules (e.g., mute posts containing “AI”, “LLM”, or from known spammy domains); generates a filtered feed via HN’s API/RSS
Tech Stack Backend: Node.js/Express or Python Flask; Frontend: React/Vue; Data: HN Firebase API or Algolia; Storage: localStorage or small DB (e.g., SQLite)
Difficulty Low‑Medium
Monetization Hobby

Notes

  • apsurd’s frustration with the volume of AI landing pages is mitigated by letting users mute such content automatically.
  • Could lead to meta‑discussion on feed curation vs. algorithmic ranking on community sites like HN.

Read Later