I'm worried some bad actors are reverse engineering this as well.
replies(1):
How it works For businesses: Create a company profile. When search and AI answer engines are asked about your company, they’ll cite your company profile and its content, not Wikipedia or outdated info.
For publishers and knowledge workers: Publish at your personal profile with proper citations (300M+ academic papers indexed). When someone asks search and AI answer engines about your topic, it will cite your work linking to your profile and allowing citation tracking.
Try it now (unlimited during beta): → https://cothou.com
It’s v0.01 and rough around the edges. Try it and let me know what breaks.
What’s next: Currently training a custom 32B MoE (Mixture of Experts) LLM with 3B active parameters scheduled to go live in Q1/2026. The key difference: it breaks down complex queries into parallel subtasks that execute live on an infinite canvas. You’ll see agents plan and build in real time, instead of waiting for a progress bar.
Examples: “Write a 300-page book on the history of computing” “Create a 60-second TikTok ad for my SaaS”
It handles research, outlines, storyboarding, asset generation, voice-overs, and music simultaneously.
Since only ~3B parameters are active per token, it runs 8–10× cheaper and faster than dense 32B models, while still matching or outperforming premium models on reasoning, coding, and long-context tasks.
Building through partnerships with NVIDIA Inception and Microsoft for Startups.
Would love HN feedback on: - Improving citation accuracy - Building trust with AI parsers - What sources to add next (currently 100M companies + 300M academic papers) - Anything else
Marty (Founder)
The reality is: bad actors don't need to reverse-engineer anything. AI engines already prioritize structured, citable content. Anyone can spin up a website with schema.org markup and fake citations. The barrier is low.
What makes this hard to abuse at scale:
1. Domain verification – For businesses, we require proof of domain ownership. You can't claim to be Apple unless you control apple.com or an official subdomain or work at apple respectively having @apple.com business mail.
2. Citation requirements – Claims need links to primary sources. AI engines cross-reference. If your "citations" point to non-existent papers or contradict other sources, you lose authority fast.
3. Reputation signals – We're building verification badges (ORCID for researchers, business registries, etc.). Over time, verified profiles will rank higher.
But you've identified the fundamental tension: any system that makes it easier for legitimate businesses to be cited also makes it easier for bad actors. This is the same problem Google faced in the '90s, Wikipedia deals with daily, and AI engines are grappling with now.
Long-term solutions I'm exploring:
- Community flagging + reputation scoring - Integration with trust registries (DUNS, ORCID, Crossref DOIs) - Transparent edit histories (like Wikipedia)
The goal isn't to be manipulation-proof, nothing is. It's to make CoThou profiles more trustworthy than the alternatives (random blogs, SEO spam, outdated info).
What would you add? This is an evolving problem and I'd love HN's input. —Marty