IdeaWell

Project ideas from Hacker News discussions.

There's a 100% Chance AI Agents Are Ruining the Internet

Original Article

Hacker News Discussion

📝 Discussion Summary (Click to expand)

Four prevalent themes in the discussion

  1. AI agents are flooding the internet with low‑quality, spam‑like content and eroding authentic human interaction
  2. “The golden era of nerd filled internet is over. Now, even nerd parts of the internet are getting filled with AI slop and marketing.” — avgDev
  3. “I stopped reading papers, journals, blog/internet posts, books, etc written after 2022 unless lots of people repost it as high quality … I basically stopped using reddit, threads and hacker news.” — SpaceManNabs

  4. “We need a hard social stigma against allowing your AI agent to contact another human being on your behalf.” — simonw

  5. Many users are retreating to offline, real‑world activities as a reaction to the degraded online experience

  6. “I’m moving more and more away from the internet … I will probably just do some more nerd stuff IRL where I can meet people.” — avgDev
  7. “My time reading older (e‑)books has skyrocketed … I recently noticed I need to stop watching youtube videos made after 2024 too.” — SpaceManNabs

  8. “I’m definitely viewing less than I did last year.” — robotmay

  9. A legal or regulatory approach is seen as necessary, since technical fixes alone won’t stop abusive AI agents

  10. “The solution is probably legal, not technological. We need a legal solution to automated spam, AI or not.” — AlienRobot
  11. “In that case the startups should be able to defend themselves against the social stigma … but here’s why we think it’s worthwhile anyway in this case.” — simonw (defending a legal‑focused stance)

  12. “If humanity's ‘solution’ to AI is to go to the government … then we’ll truly have fallen into a poetic trap.” — jMyles (highlighting the reliance on government/law)

  13. Efforts to block bots (captchas, verification hoops, Cloudflare challenges) are creating significant friction and hurting legitimate users

  14. “Now, it takes 8 seconds to first pass a 'checking if you're a bot' page or worse - making you click on pictures of crosswalks.” — ks2048
  15. “I absolutely loathe the human verification hoops I have to jump through now, more than anything else mentioned in this thread.” — distortionfield
  16. “Cloudflare has gotten out of hand for me lately … I'll just get straight up blocked from normal browsing.” — rurp
  17. “I can't go a day without being blocked from a site because it thinks I'm a bot.” — dlcarrier

🚀 Project Ideas

AgentID

Summary

  • Provides a lightweight, cryptographically signed header for AI agents to self-identify as non-human, enabling platforms to label or filter AI traffic.
  • Core value: reduces unwanted AI spam and lets users enforce social stigma or policies based on agent transparency.

Details

Key Value
Target Audience Developers building AI agents, platform operators (APIs, forums, SaaS)
Core Feature Standardized attestation token (JWT) that agents include in requests; verification service and reputation ledger
Tech Stack Node.js library, Go verification service, optional IPFS or blockchain for reputation; OpenID Connect compatible
Difficulty Medium
Monetization Revenue-ready: tiered SaaS for verification API (free tier, paid for high volume)

Notes

  • N_A_T_E asked: “Is there any solution? Can sites, apis or interfaces require AI agents to identify themselves as non-humans?” – AgentID gives that solution.
  • simonw noted stigma only works if agents are identifiable; AgentID makes stigma enforceable.
  • Could become a de‑facto standard like robots.txt for agents, sparking discussion on responsible AI deployment.

HumanRank

Summary

  • A browser extension and search API that boosts content likely created by humans and demotes AI‑generated slop using linguistic heuristics and user feedback.
  • Core value: restores signal‑to‑noise ratio for readers tired of AI‑filled feeds.

Details

Key Value
Target Audience Power users, researchers, hobbyists who consume articles, papers, forums
Core Feature On‑the‑fly scoring of pages (perplexity, em‑dash ratio, burstiness) + community voting to adjust weights
Tech Stack Chrome/Firefox extension (TypeScript), background worker using TensorFlow.js or HuggingFace transformers, optional backend for model updates
Difficulty Medium
Monetization Hobby

Notes

  • SpaceManNabs said he stopped reading papers unless “lots of people repost it as high quality” – HumanRank surfaces those human‑vetted items.
  • soupfordummies called for a disclosure method; HumanRank works without requiring disclosure, just detects AI‑like patterns.
  • Practical utility: users report less time wasted on AI junk and more chance to find genuine content, a frequent HN gripe.

SafeSpeak

Summary

  • A virtual receptionist service for small businesses that uses an AI voice agent but always discloses its non‑human nature and offers instant handoff to a human.
  • Core value: lets contractors capture leads without scaring customers away with hidden AI.

Details

Key Value
Target Audience Contractors, tradespeople, local service providers (plumbers, electricians, concreters)
Core Feature AI greeting that states “I’m an AI assistant helping [Business]; I can answer questions or connect you to a human”, integrated with calendar/FAQ, with easy transfer
Tech Stack Twilio Programmable Voice, Whisper for speech‑to‑text, GPT‑4o for response, simple web dashboard; optionally VAPI
Difficulty Medium
Monetization Revenue-ready: $15 per line per month + per‑minute usage

Notes

  • cube00 warned: “Your own authentic voice answering a customer query directly is one of the few advantages a small business has over the big guys.” SafeSpeak preserves that advantage by being transparent and offering human fallback.
  • avgDev described the pain of contractor hunting – SafeSpeak reduces missed calls while being transparent about AI use.
  • Addresses ethical AI use in local commerce and could become a selling point (“We’re AI‑transparent”), a topic often debated on HN.

HumanCheck

Summary

  • A privacy‑first, accessible verification widget that replaces frustrating image CAPTCHAs with quick, human‑friendly challenges (e.g., simple gesture, audio cue, or trivia) that are easy for humans but hard for bots.
  • Core value: eliminates UX pain while still guarding against automated abuse.

Details

Key Value
Target Audience Website owners, SaaS platforms, anyone using CAPTCHA today
Core Feature Embeddable JS widget offering a choice of micro‑tasks (drag‑slider, spoken number, picture‑tap) with adaptive difficulty; includes accessibility fallbacks
Tech Stack Vanilla JS widget, optional backend node service for challenge generation and verification; can use WebAuthn for device‑based proof
Difficulty Low
Monetization Hobby (open source) – optional hosted version for enterprises

Notes

  • distortionfield complained: “I absolutely loathe the human verification hoops I have to jump through now” – HumanCheck removes that loathing.
  • ks2048 noted the annoyance of “making you click on pictures of crosswalks”; HumanCheck offers alternatives.
  • Potential for discussion: improves accessibility and reduces friction, a recurring theme in HN threads about bot defenses.

Read Later