Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

You're absolutely right, and I apologize for the confusion.

The demo is a single-session system (designed for hotel lobby kiosks where one person uses it at a time). If someone else is talking to the avatar when you visit, you won't see a QR code.

Here's a direct link you can try: https://ps.neoclerks.com:8451/player.html?ss=f2ee51d8-2ea3-4...

If that's busy too, I'm happy to schedule a 15-min private demo. Email: EchenDeligani@gmail.com

This single-user limitation is product design (kiosk model) not technical - production handles 10-50 concurrent users per GPU. I should have explained this upfront and provided multiple demo instances for HN traffic.

My mistake - learning as I go here.



Not trying to "gotcha" here, but are you using AI to converse?


They might be using Claude as a translation app and not realizing that the AI style stands out in English.


when AI is talking better than why shouldn't I use it?


People will think it's fake. Also, it hard to know what is real and what is AI hallucination.

Let's try a weird question: Can you reply in Persian which are your 2 favorite icecream flavors and why?


That's exactly the use case I would go too (since I'm not native English) and while we have a technology now for this, people will straight away reject because LLM. Funny how we always wanted to connect the world and know it is just not compatible with the situation.


> You're absolutely right

Is this botspam?


This one also comes up from foreign language speakers using ChatGPT or Claude to translate their point, and then not trimming the GPTisms.


Yeah, I'm using Claude to help me write responses. English is my second language and I'm nervous as hell about this thread.

I'm reading the comments, figuring out what to say, then using AI to help phrase it properly. The thoughts are mine but the wording is AI-assisted.

Should've been upfront about this from the start. My bad.

The product is real (test it at neoclerks.com/en if you can access it), but I get if the AI responses kill the credibility.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: