Brilliant Labs' frame AI glasses | Hacker News Reader

Brilliant Labs' frame AI glasses

brilliant.xyz

101 pointsgeox54 comments

So, I was barely able to view the website (I'm not sure if that's just me, or if that is how the site is supposed to be designed, but from the youtube linked by @Topfi, is this a good use case for the glasses? Why would I want to read that information? Meta glasses can do this and speak the info to me, which seems a nicer UX.

There are definitely things that would be better in the visual field, directions is the obvious one, but that's been seen so many times. Seeing as we all QR code menus at restaurants now, how about scanning and showing a menu. Sure it isn't AI, but that's not the point of the glasses.

I find these devices, Humane and Brilliant, focus too much on the AI component when the magic is the hardware they've developed. AI is and will be everywhere. It can run everywhere. I feel we're already at the point where "it has AI" is as valuable as "we've built an iphone app that..."

Rabbit is the exception here, which is why they standout. The AI is the compelling part of their offering, and the hardware is designed to limit the availability of the AI giving the early adopters the opportunity to train the AI on the tasks it will be able to complete for the mainstream, when the market gets there.

I am cautiously excited for these, because they're almost normal-looking enough to be useful, and because you can just use bluetooth & a python library to drive them, you're not tied to some godawful app... they might be just what I'm looking for in a wearable computer I/O device.

However, based on my previous experiments in wearable displays, I'm also extremely dubious about their performance outdoors -- I'd want to hear from somebody outside the company who's worn them outside before I bought a pair.

More normal looking than the Meta/RayBan glasses?

No, but they have a display and the Meta glasses don't, so I don't really compare them.

Same, I have some guarded optimism about the cool factor of a hackable wearable display. Less enthusiastic about the camera and the AI integration stuff, and not sure I love the thought of having to charge my glasses. But since I wear glasses anyway, seems kinda neat. The $350 price tag feels sort of reasonable for the potential.

Still, I think I'd need to see some serious UX dedication here, because it's the kind of thing where if it's not basically perfect it's going to be incredibly annoying. I can imagine a ton of utility for this kind of thing but it might not be easy to capture that usefulness and make it feel seamless.

Neat thing though!

Nothing that has AI in it in 2024 is going to be anywhere remotely as useful as it seems.

And even if it is, it's not going to be portable from a battery and remote access point. If it's on the device, the models going to either be garbage or drain the battery. If it's not on device, the signal and connection will drain the battery.

Just look at what Apple found "acceptable use" and they don't even have any killer app atm.

technologies progress like evolution: they don't have morals, don't consider binary existence, and proceed whether or not we choose them. it's a part of culture.

we barely get cell phones that last all day with normal use.

Regardless of all the tech, the battery is still the weak point of any "slim" offering. Even if you can get perfect vision, you arn't gonna get something with a practical use.

It’s concerning there isn’t a single image of them being worn by a human.

Yea it’s crazy that there is a photo of the three engineers (who all wear glasses) but no one is wearing their own product.

To fit the necessary tech in, the glasses are scaled up 2x.

Maybe you kinda look like Harry Potter if you wear them.

Or Waldo (wherever he is).

He is working on finding himself

interesting to look at their site from a couple years ago. no mention of AI, they are simply "Open source eyewear."

https://web.archive.org/web/20220809080848/https://www.brill...

If you scroll to the bottom of the linked page, they have a link to the Monocle product

"Scrolling to the bottom of the page" is easier said than done, since it's doing that annoying scrollbar hijack thing that's apparently still in vogue.

The archive.org link? It scrolls normally, though part of the background does not scroll.

Ohh the Monocle is cute: https://uploads-ssl.webflow.com/623cc6cc56889b045032bfc1/62a...

if Google Glass had come out later you can be guaranteed that it would've been billed similarly.

There is a formula here : Make a product with a semi-useless privacy-intrusive camera, when the audience balks at having a camera on it you say "it's AI!" -- suddenly the people in front of the camera seem to stop caring that they have a new angle of the panopticon to avoid.

It's the damnedest thing. my gut feeling is that trick would have worked to make the Glass palatable -- similar tricks worked for the current iteration of VR sets : 'It's okay that the device has six cameras pointed at you recording to god knows where, they're used for tracking.'

the trick being that the tracking included is unlikely to only be your hands/posture/VR-significant things if it stands to produce company X or Y a revenue stream.

True, the utility of a camera draws attention away from any nefarious possibilities. I also think our relationship with cameras is different from when google glasses came out. My seven year old niece will film herself dancing using her mums phone and even though its just recorded to the phone she'll intro the video 'hi everyone', probably from the content she's exposed to, tik tok etc. 'Audience by default'

Also, it was Google. I still would not wear something like that if it came from Google (or Meta, or any number of companies that make their money off our data).

Does anyone have any interesting Monocle use cases to report? I ordered one and tried hacking around with it, found very little in terms of durable utility or even hackability for interesting computer vision/AR applications, so would be curious to hear if anyone did better.

That's not to knock the creators btw, I think they're working right at the edge of some incredibly challenging hardware constraints, just that maybe the tech is still immature for those who want to apply it, as opposed to those who want to hack on the tech stack itself.

Also if anyone wants my Monocle to hack on themselves, I'd be happy to send it off for the cost of shipping!

Edit: Claimed

I've been real curious about it, how do I contact you?

edit: I thought it might be suitable for the wearable computer I've been tinkering with at home... resolution is better than the display I've got now, although their API for displaying things is really primitive.

I played around with the monocle - it was interesting but very limited.

https://youtu.be/lOoptAtZMjk

my criticism of this is the same as that silly AI pin, or the thing from CES everybody was excited about: just make an app. if your product is camera+ai+outputs, nothing about that needs dedicated hardware.

if your app is so good i want to wear it on my face all day, then that's awesome, and i'll buy the glasses. but i don't want to have to spend hundreds of dollars on a pair of glasses just to find out that the software is underwhelming.

Not sure I follow. Same argument could have be made for smart phones (just make a PC app). This is a new app platform. It may or may not work, but it’s not a single purpose device.

Except that smart phones already had a concrete use case beforehand (web browsing/checking emails on the go, instant messaging + phone calls) which only got better with advancements in technology (ie, iPhone). This is putting the cart before the horse. There are no great applications on phones that would be improved with new hardware nor are there unique applications that can only be done on the hardware without the phone.

yes, exactly: the argument for smartphones was obvious: "i want to take this app that i use on my computer, and take it everywhere with me in my pocket".

that same argument doesn't exist for these new AI devices. there's no reason these point-a-camera-and-get-an-ai-response apps can't exist on phones, or computers, but they mostly don't. so who's calling for putting those apps that don't exist on your face?

While they do have a previous project in the Monocle, I am very skeptical that these glasses will be shipped anytime close to the current date (the 15th of April), considering they have only one video [0] that shows these glasses outside of renders. Add to that, the singular piece of content showcasing these glasses in the real world in no way inspires confidence; there is no way to tell whether this is merely a simulation, recording through the glasses, etc. Add a bit of shacking to a video of a cake and some text overlay, cut to a few seconds of you wearing some glasses beforehand and the result would look identical.

If the first batch of preorders (not even just reviewer samples) are going to be shipped in mid-April, production would have to ramp up some time soon, and the final hardware should be fixed, so why not show us that in action? Also, I am very much not an expert, but the camera sensor and pinhole in front of it seem too small for what would be demanded of the two, and I struggle to see how such a small MicroOLED panel could yield the massive high-resolution images [1] and nearly full lens coverage they showcase in some of their images, especially considering my previous experience with competitors [2]. Even more astonishingly, they claim an IPD range of 58-72mm, which also raises my skepticism quite a bit.

Also, they don't seem to limit the first batch to a single color, even though that is generally recommended to limit manufacturing complexities during batch releases, and they take orders for prescription lenses. Producing three distinct colors with the many variations of prescriptions and getting that shipped by Mid-April, makes this all even harder to swallow, especially at such a low price.

Lastly, their "Specifications" [3] aren't really what I'd normally consider as such. I am a firm believer that products of this type can be more than the sum of their parts, and a certain "je ne sais quoi" can often not be read from merely looking at specs, but still, I'd feel much better if something tangible was listed.

If this product (specifically the part about covering the majority of the lens with display content [1]) ends up real or even anywhere close to that, I'll be the first to purchase it, even if the final price ends up significantly higher. Heck, I'd settle for a new release of Norths Focals with their tailor-made (as in uniquely fitted to one wearer) design, small sweet-spot and low resolution.

But as it stands, I can't really say I have a lot of confidence in them delivering. I hope that this is a case of ambition and overly optimistic enthusiasts that will end up pulling through (after some delays) for the sake of anyone who bought in. If that is the case, I'd advise the creators, both for their personal sake and to help their preorder customers retain confidence, to release as much uncut, directly captured footage as possible. It doesn't have to (in fact, shouldn't) be overly fancy. Considering their focus on being open-source and the people that tends to attract, I find it surprising that this isn't their mode-of-operation in the first place.

Lastly, I have found a picture of what appears to be an early prototype [4], which has a far smaller display section, much more in lign with what seems possible using the design they showcased in that exploded view, as far as my understanding of how this tech would work goes [5].

Should they execute on what they promise with those renders in terms of display area and resolution, I'll be more than happy eating a large helping of crow.

[0] https://www.youtube.com/watch?v=eXXTwv3sXto

[1] https://imgur.com/TtunqYQ

[2] Personal experience: XREAL Air 2 Pro (MicroOLED), Viture One (MicroOLED), Meta Quest 2 (LCD), Meta Quest 3 (LCD); I only retained the Quests after testing

[3] https://brilliant.xyz/products/frame

[4] https://www.reddit.com/media?url=https%3A%2F%2Fpreview.redd....

[5] https://news.ycombinator.com/item?id=39321973

That... is not how lenses work, right? They are showing a totally flat glass surface that close to your eyes? I can't imagine your eyes going to be able to focus on the entire width of your FoV of text and images of that size... without it at least giving you a headache

It looks interesting but the prescription option seems kinda useless. I have less then 0.5 on both eyes. But it doesn't ask for my eye distance or astigmatism. So it's useless for me?

More discussion: https://news.ycombinator.com/item?id=39310334

No where near enough info here to actually stump up so much cash - but I do like the idea of AI glasses. I already have to put them on my face and carry the around and pay a lot of money for them, so this could be a good way in.

If Apple sort out Siri though my Apple Watch is gonna be suddenly very useful

The website says these glasses have all-day battery life. Is that with the use of a portable charge? If so, I'm wondering what the battery life is without the charger.

Really excited for this after Google discoutinued Google Glass, at least for me use this thing to play some Janpanese games are awesome, perfect lol.

Does it support the myopia lenses?

Why such a dorky Benjamin Button rounded lens frame design? Do they want us to look like Pixar's old man, Geri? Lol.

https://www.oklahoman.com/gcdn/authoring/2008/12/22/NOKL/ghn...

https://upload.wikimedia.org/wikipedia/en/0/08/Geri%27s_Game...

Hopefully these brilliant folks will wake up and modernize the design and ditch the trashy PayPal tramp stamp on the site footer.. you know, before they run out of funding.

(p.s. sorry if this was a dickhead post, I really do hope they will improve it so I can then purchase via literally any other method)

Given that the USB-C charging adapter makes them look like those joke shop disguise glasses, I can only assume that this is deliberate. At least I respect someone who makes no effort to look cool over someone who tries really hard to look cool and fails.

Not sure why anyone would use “nutrition search” it looks cool but everyday life no one is gonna use it other than to check it out

Magic Leap vibes. I predict nothing will come out of this.

Products like these are the future. Not the apple vision pro. These are the new iPhones whereas the apple vision pro is the new blackberry.

wow a site loading progress bar, that's a blast from the past.

Is it one display or two?

I like the website

This looks like a scam. We don’t have fully see through screens yet.

this is false

https://www.youtube.com/watch?v=2BZy3OJeAbs

in any case, I doubt that's what's being used here though

The exploded view [0] showcases them placing a very small MicroOLED panel at the top of the frame, whose light then appears to get projected into your eye using something they call an "AR Lens".

In principle, this is roughly comparable with the approach Viture, XREAL, Nreal, etc. have taken. Each places a 1080p MicroOLED panel at the top of a prism that bounces the image into your eye, covering the top third of your vision. That's one of their main disadvantages and more than likely why neither Meta nor Apple utilize this tech, it really works best if you combine very tall nose bridges and then push it as far forward towards the nostriles as you can, in hopes of getting the top section with that prism roughly at eye level. Again, Brilliant Labs claim an incredibly wide-range of FOV and nearly full lens coverage from the center, not the top.

Also, the prisms used be the competition require far more phisical depth and the MicroOLED panels in those seem (very rough estimate using the exploded view rendering as a reference) two times larger in area size, so I have a hard time understanding how Brilliant Labs could manage such full screen coverage at such high quality, with such a thin "AR Lens" and what likely would be a lower resolution panel. In fact, the panel size looks incredibly close to the panel size in the render of their Monocle [1], which is 640x480 and projected using a similar prism solution into the lower third, not across the entire lens, which would be far more realistic and believable.

I'd be a bit more trusting if their images, representing what the display is supposed to look like, leaned towards the North Focal direction of very simple, low-res graphics. That would indicate a massive success in designing their lens solution and would make sense with a smaller MicroOLED panel. But high-resolution, full lens coverage using something fairly slim; that's a lot of sudden innovation that may deserve a more in-depth presentation, ideally with some independent journalists to verify.

Again, I'm mainly skeptical because I'd really love for this to be real.

[0] https://imgur.com/MnOW2Rr

[1] https://docs.brilliant.xyz/monocle/images/monocle-exploded-v...

I think what you decribed is birdbath. RayNeo X2 and INMO (Air2 or Go) use waveguide display as the "see through screens" you described. It doesn't seem to be radically better than or even as good as RayNeo X2 Lite, scheduled to release in 2024, so it's not that unreal. (Or if they are targeting single color display then INMO Go is a closer comparison)

> Brilliant Labs claim an incredibly wide-range of FOV

The website says ~20 degree diagonal FOV, isn't that really low? I thought XReal/Viture/etc had around 40 FOV.

Sorry, I made a mistake, I accidentally typed FOV, when I meant IPD. The IPD they clame for this product has a range of 58-72mm, which is very wide when comparing the competition. They made no claim of FOV, from what I've seen, for the glasses, ~20 degrees was for the Monocle.

Oh okay, that makes sense. If you go to preorder, and then FAQ->specifications, it does say ~20 deg for the frames too.

Huh? We've had commercial see-through displays for over a decade, of various qualities and features.

Irrelevant here anyways since the schematic on the landing page clearly shows that it's a small OLED in the top of the frame itself either angled directly at an eye or more likely using specially cut glass to make the image visible to you.

Considering this company has already shipped a product, with a see through display, I think you’re probably wrong.

Reply on news.ycombinator.com