Recent submissions

Comment · 1 year ago
We are announcing the development of our LLM-powered smart watch, called Rist. We're developing an open, cloud-based LLM operating system that makes it easy for users to connect their own LLM workflow endpoints, enabling custom tool usage and assistant modifications. Users can even fully replace the provided cloud OS with their own custom-built assistant or OS if they’d like.

It is platform agnostic, meaning no matter where you build your LLM workflow or agent, or which models it uses, you can easily connect it to the watch by providing an endpoint to your workflow, making sharing workflows among users easy and seamless.

As a fully standalone device, Rist features LTE connectivity and manages all standard smartphone functionality through voice and text input. Phone calls, navigation, media streaming (currently audio only), email, etc. are all handled by specialized LLMs. If any workflow doesn't meet your needs, alternative options will likely be available from other developers, or you can integrate your own custom solution.

It is our goal for the Rist to be a highly productive, less distracting alternative to smart phones for on-the-go use and digital minimalists. We leverage LLMs to maximize user productivity for tasks that would normally be performed on the phone, while access to social media or modern electronic entertainment options are naturally limited by Rist’s form factor.

We looked at a lot of previous devices like the Humane AI Pin, Rabbit R1, and others, and we've opted for a simpler, more open, and developer-friendly approach. Rist provides access to the rich ecosystem of LLM workflows that developers have built across various platforms, allowing them to integrate applications without navigating a lengthy approval process.

This first version has a minimalist interface with text and voice input, with optimized battery life designed to last for days. The watch supports displaying images, text, and playing voice or streamed audio content. We'll provide comprehensive documentation for developers on managing these capabilities through their endpoints.

We are a small team of talented engineers working to keep costs down as much as possible so that we can meet our price point goal of under $100 per unit with a few color options.

Join our waitlist for updates and early access: aris.chat/rist-waitlist

Comment · 1 year ago
You're right. Parents can google it and explain it to their kid. Parents can find the answer in an encyclopedia, find a book about it, or find a youtube video for them. There are lots of options. I hope we didn't offend you.

As parents ourselves, we think it's helpful for kids to have independent access to the world's knowledge. We've found it helps them feel they can learn as far as their curiosity takes them instead of waiting for library day or waiting for us to look things up for them. Statistics show that most parents just end up letting their kids use Google independently, which has led to half of children today stumbling on adult images and videos by 12. For some reason, I don't see a lot of people criticizing that. Perhaps, because they don't market to kids. Kids just use it. That's probably what will happen with AI tools that don't specifically market to kids. They will have adult content, kids will get exposed, but they won't be blamed because they didn't market to kids.

We understand all parents have different opinions on these things. We basically just built the tool we want for our kids, and we want to give interested parents an extra option if they want it. One where it is impossible to find adult images, video, links, or advertisements.

Comment · 1 year ago
We have added a much thicker filter layer for the moderation, so it should be working much better now.
Comment · 1 year ago
This has been fixed now. Thank you for pointing it out.
Comment · 1 year ago
Thank you for this. We are fixing this now.
Comment · 1 year ago
For questions their parents know the answers to, absolutely. That is the best idea. But the alternative to an LLM chatbot isn't their parents, it's Google. And Google has been showing kids pornography for decades. So, why not give kids access to information that does not also have access to pornography?
Comment · 1 year ago
We will think more about that. Maybe we can allow parents to comment, rate, or correct responses in the child's logs. Thank you for the feedback and suggestion.
Comment · 1 year ago
I use LLMs to answer questions every day, and I've found it to be healthier than Google where the average age for kids to stumble across adult content is 12.
Comment · 1 year ago
It's short for Aristotle, so the pronunciation may be different.
Comment · 1 year ago
I have heard of books, but my kids prefer to have a large percentage of the knowledge of books on their wrist, such that if they see a clam at the beach and want to know about it, we don't need to wait until library day on Tuesday.
Comment · 1 year ago
Thank you, this is an easy fix. We will add this change ASAP.
Comment · 1 year ago
That is good. Our goal to be competitive is to be cross-device and to try to give parents a level of control over information that they've not had before.
Comment · 1 year ago
Wonderful feedback, thank you. We are a movies only and only on the weekends family as well.

- We log the questions asked in each child account for parents to view in the parent account, but the questions are only logged when they're signed in.

- Regarding giving the answer rather than fostering the process of discovering the answer, this is something we've gone back and forth on. I think we will add a setting for this in the parent's settings for each child account. We've talked to parents who want one or the other, so we will work on adding it as an option to just give them answers right away or have it start those teaching moments.

We appreciate these suggestions, thanks!

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton