I built an anonymous Q&A space where **only humans can reply**, no AI — 1 month real‑world traction update
Hey IH,
Tired of AI‑generated generic replies everywhere online, I built a side‑project: an anonymous question‑and‑answer community with one hard rule: no AI responses. Every answer comes from real human visitors.
No login, no email, no phone number required. Users post vulnerable questions, strangers respond with genuine human perspectives.
The site auto‑serves different language pools: English‑speaking visitors see English‑only conversations, Chinese‑speaking visitors get a separate Chinese‑language feed, they do not mix with each other. Every question and answer has a built‑in report button for moderation.
It’s been live for a little over one month. Raw unpolished numbers so far:
- ~1100 unique user sessions (uses anonymous recovery token instead of accounts)
- 458 public questions submitted
- 1049 human replies generated by community
Some days we get steady organic questions, other days are quiet. I’m a solo developer running this on Vercel. There is no monetization right now.
Early observation: It’s surprisingly hard to get real‑human engagement. Many anonymous platforms get lots of questions but very few thoughtful replies. The karma‑like reward system I implemented helps encourage people to write answers, and it actually works better than I expected.
I tried posting to Reddit with a brand‑new account, and my links kept getting auto‑removed by spam filters. That’s why I’m sharing here.
I would love feedback from this community. What would you improve? What risks do you see for a human‑only anonymous UGC space?
Check it out: https://askamonk.online/
Thanks.
The reply rate is the interesting signal here.
Have you seen whether rewards increase replies independently of question quality?
Measurement boundary between volume and signal quality. Questions submitted (458) is a raw count - answers generated (1049) looks bigger until you ask: how many questions never get answered? The karma system doesn't just encourage replies, it's a measurement device making visible which answers people trust. Separating English/Chinese pools splits the measurement domain - same question gets answered differently in each language ecosystem. That's not just moderation, that's measurement infrastructure deciding which problems get solved by which community. The hard problem isn't engagement, it's signal fidelity - distinguishing between "this answer is popular" and "this answer is actually helpful."
Really Thanks for the deep insight.
You’re totally right—raw numbers don't mean much if questions go unanswered, and separating "popular answers" from "actually helpful answers" is hard.
I'm actually working on these exact issues now:
Fixing the unanswered questions gap.
Improving the Karma system so real solutions rank higher.
Fine-tuning the moderation for both English and Chinese pools.
Really appreciate you taking the time to break this down!