IdeaWell

Project ideas from Hacker News discussions.

Erdosproblems.com Succumbs to the AI Onslaught

Original Article

Hacker News Discussion

📝 Discussion Summary (Click to expand)

1. AI‑generated proofs are eroding the site’s original purpose
- “The main way that people publicly interact with the site now is to advertise their AI‑generated proofs, often without any attempt to explain them, but as a way to record a (increasingly meaningless) priority claim.” — pfdietz
- “I find it sad that I can't use the site as a tracker for solved problems anymore.” — bananaflag
- “People posting AI‑generated proof to claim credits are the scourge on mankind discoveries.” — Qiu_Zhanxuan

2. AI can be useful if separated from human‑focused mathematics
- “I believe that websites with this function should exist … places where people can record AI‑generated proofs … to save others wasting their tokens generating the same proof.” — nemomarx
- “The excitement from solving these problems should be considered a good thing. More people engaging and having fun with math is positive.” — charcircuit
- “I especially like that he’s allowing AI if it generates a better learning proof.” — jwpapi

3. Broader ethical and societal concerns about AI’s impact
- “The benefits of AI accrue disproportionately to those most deficient of scruples, while its costs and harms fall upon the selfless and pleasant …” — fwlr
- “The fracking or strip mining of mathematics is repulsive.” — afsg‑qsgf
- “Obviously those that benefit from AI have no ethics and morals.” — 12997

🚀 Project Ideas

Generating project ideas…

ProofVerify

Summary

  • A middleware layer that flags and separates AI-generated proofs from human-verified solutions on problem‑tracking sites, requiring explanations for AI submissions.
  • Core value: restores the site’s usefulness as a tracker of genuine mathematical progress while still allowing AI contributions to be visible and searchable.

Details

Key Value
Target Audience Moderators and users of open‑problem trackers (e.g., Erdős problem list, Polymath projects)
Core Feature Automatic detection of AI‑generated proof submissions, mandatory explanation field, and togglable filters to view human vs AI proofs
Tech Stack Python/FastAPI backend, PostgreSQL, optionally HuggingFace transformers for AI‑text detection, React frontend
Difficulty Medium
Monetization Revenue-ready: SaaS subscription for site owners (tiered by monthly active users)

Notes

  • HN commenters lamented “I can't use the site as a tracker for solved problems anymore” (bananaflag) and wanted a separation between AI and human contributions.
  • Provides a practical utility by reducing noise and enabling reliable problem‑status tracking, fostering discussion on proof quality.

AIProofHub

Summary

  • A searchable repository dedicated to storing AI‑generated formal proofs (Lean, Coq, Isabelle) with deduplication, versioning, and citation metadata.
  • Core value: prevents wasteful recomputation of the same AI proof and gives researchers a citable source for AI‑derived results.

Details

Key Value
Target Audience Researchers, AI‑assisted theorem provers, formal methods engineers
Core Feature Upload, search, and download AI‑generated proofs with similarity hashing to avoid duplicates, plus export to common proof‑assistant formats
Tech Stack Node.js/Express server, MongoDB for proof metadata, IPFS or S3 for storage, Lean/Coq CLI integrations, Vue.js frontend
Difficulty High
Monetization Revenue-ready: Pay‑per‑download or institutional licensing (API access fees)

Notes

  • Commenter nemomarx argued for “a site for agents to read … to save others wasting their tokens generating the same proof” – exactly the need AIProofHub addresses.
  • Encourages scholarly reuse of AI proofs, reduces compute waste, and can spark discussions on proof‑checking standards.

ProblemTracker+

Summary

  • An open‑problem tracker that emphasizes human understanding via threaded discussions, incremental progress annotations, and optional AI‑assisted hints that are clearly labeled.
  • Core value: revitalizes the collaborative, explanatory spirit of problem solving while still leveraging AI as a tool, not a claim‑generator.

Details

Key Value
Target Audience Mathematicians, hobbyists, educators seeking a community‑driven problem‑solving platform
Core Feature Problem pages with timeline of human contributions, discussion threads, ability to mark steps as “human‑verified”, “AI‑suggested”, and to attach explanatory narratives
Tech Stack Elixir/Phoenix (real‑time updates), PostgreSQL, GraphQL API, React with Redux, optional integration with OpenAI API for hints
Difficulty Medium
Monetization Hobby (open‑source, optional donations via Open Collective)

Notes

  • Users like pfdeitz expressed sadness that the site’s purpose was lost to AI spam; ProblemTracker+ returns focus to human‑curated explanations.
  • Provides a venue for practical utility: collaborative problem solving, teaching, and generating discussion that could lead to new research directions.

Read Later