Brilliant Labs Frame Review
Can this pair of smart glasses fill the void?
The Brilliant Labs Frame is one of the hardest products to evaluate that I’ve ever had a chance to review. The Singapore-based company classifies the Frame as a “pocket-sized AR device for imaginative hackers” and that itself kind of sums up what the experience of this product actually is. I definitely don’t have the skillset of a hacker, nor am I savvy enough to develop applications. I am the definition of a general consumer when it comes to smart glasses. After spending a couple of months desperately trying to utilize the Frame into my daily life, I have a hard time even classifying it as a consumer product. And that’s a bummer seeing as I was extremely excited at the prospect of an AI powered pair of glasses.
Purchase Price: $349.99
Resetting Credits
Each interaction with the Frame uses Monthly Credit.
I’m going to start off by at least giving Brilliant Labs the pat on the back that they deserve. The Frame exists. You can buy it and it will show up on your door. In the age of crowdsourcing, consumers have become all too used to being let down with unfulfillable delivery windows, delays, and even cancellations of hyped products that ended up being vaporware. That isn’t the case with the Frame. Like the Monocle before it, Brilliant Labs made an announcement, declared a release date that was only a mere few months after, and remarkably did release the product within a reasonable range of time from that date.
I was not sent a review unit. I pre-ordered it just like any other consumer and I received the package in our office safe and sound. I’m 100% sure Brilliant Labs had never even heard of our publication. So yes this is a legitimate pair of smart glasses that anyone can purchase and actually get on your face. There’s not many of them on the market right now. Trust me, I’ve been an advocate of this wearable space for ages. Anyone remember North Focals?
What Kind of Smart?
A micro OLED screen projects NOA’s interactions
Now onto the nitty gritty that I hope will answer most of the questions out there regarding the Frame. Structurally, the product is very simple. It’s a pair of plastic glasses that houses a camera on the brim of the nose, and a micro OLED screen embedded onto the glass of the right eye that projects text from the assistant the company named NOA. There’s also a microphone on the Frame that can hear verbal inquiries from the wearer, but no speaker like on some other smart glasses that’s primary function is to play music.
The Frame’s primary and essentially only function is to give wearers access to OpenAI ChatGPT to interact within the confines of your right eye. By tapping on the side of the frame, a prompt asking for instructions will appear in the right eye. Speak to it with a prompt, and NOA will attempt to supply you with the information you need.
In theory, you could be sitting across from a person who only speaks Japanese, while you only speak English, and have a translated transcription of the dialogue in your eye. Or say you’re out at the store and there’s an item you want to learn more about, the Frame’s camera can see what you’re seeing and relay the information you want to know about without needing to manually search for it on the internet from your mobile phone.
The NOA app has a log of all your interactions with the Frame
You can see how a pair of glasses like the Frame would be attractive to someone looking to explore the world with the help of technology. Unfortunately, the actual product experience doesn’t come close to emulating the fantasy many of us have for this technology. This ideal world starts to collapse at the core of how the Frame functions.
For NOA to operate in its base form, Brilliant Labs utilizes ChatGPT as essentially the brains of the operation. In fact, seeing as this particular pair of smart glasses can only contribute to your wearing experience with AI information, it doesn’t do anything else other than make a fashion statement without ChatGPT. A sticking point to the mainstream progression of AI adoption that isn’t talked about enough by reviewers of new tech is the cost for the processing power? AI processing of all kinds takes a lot of resources that the hardware you’re buying doesn’t cover. The Brilliant Labs Frame may cost only $349 to acquire the hardware, but in order to use the basic core function of the device, credits are required for every time you tap to talk to NOA.
These inquiry queues range in how many credits they take up upon request, but owners of the Frame have only 2000 credits to utilize each month without the need to pay. That sounds like a lot, but from my experience with the glasses, a single tap cost me around 3-15 credits or so. In less than 30 minutes, I can easily rack up hundreds of used credits just merely attempting to get the Frame to supply me with the proper information I wanted.
There are apparently different plans that you should be able to subscribe to in order to increase the monthly credit limit. As of the publishing of this review, there doesn’t seem to be a way to acquire those levels through the NOA Frame app. I say as of the publishing of this review because a lot of things were and are not ready for consumers to get their hands on when this pair of glasses arrived at our door steps.
Not REady TO Land
Although Brilliant Labs advertises the Frame to work with both Android and iOS, there was only a downloadable app on Apple’s ecosystem for the first month or so of my review period with the glasses. To make matters more difficult, the consumer package the Frame ships with comes with absolutely no physical instructions. There’s a piece of paper with a QR code that leads to a website page with minimal amount of setup information.
Let me tell you how frustrating it was to simply pair the Frame to my iPad. Initially, the links to download the app on the app stores were incorrect as they brought users to the Monocle app which is similar to the Frame’s app, but not compatible with this device. It wasn’t until about when I was originally going to publish my full review, nearly two months later that I noticed the Android app had appeared in the Play Store. I thus delayed the publishing of this article and re-paired my Frame with my Android phone.
I had kept my pair linked to my iPad in the studio up until the Android app had released as I currently am not reviewing or using an iPhone. As we review products based on consumer applications (since we are all consumers), someone with an Android phone most likely won’t have a spare iPhone laying around unless they were a developer.
Anyway, my point is that it took a bit of time, but I noticed Brilliant Labs released a couple of YouTube Shorts that were linked on the app that troubleshooted the setup process. This would have been immensely helpful during the first month of ownership as I was literally pulling out my hair in frustration.
Design
To get a better understanding at just how rudimentary the Frame is, there’s no physical buttons on the entire device. You can’t physically turn the pair of glasses ON or OFF. In order to charge the glasses, Brilliant Labs designed an orange colored pin connector that is shaped like a nose to connect with the bridge of the Frame. Cleverly named Mister Power, when it is snapped onto the Frame, the device will prevent inquiries from occurring unwanted. It’s basically a physical stopper to put the Frame to sleep.
The problem with this design is that while Mister Power is a cute implementation, it also impedes another item that occupies that spot when the Frame is on your face. Included in the purchase are two magnetically attachable nose pads of differing sizes. This piece is crucial to framing the Frame to the right eye as it keeps the device resting at the proper height. That means that every time I take off the Frame, I have to unclip the magnet nose pad and snap in the pin lock. That makes it two small pieces that need to be carried around and interchanged frequently in daily use. That’s a lot of potential for either losing a piece or damaging the pins. I would have just liked to see one small physical button on the body to shut the Frame off.
Mr. Power with a sim remover tool
On the bottom of Mister Power is a USB-C port which you will then need to plug in to charge the device. It takes a few hours to fully charge the Frame, but I’m also annoyed that there’s no easier way to monitor the battery percentage. The app is barebones and that’s usually where I would expect to at least have a reading of how much battery life is left. Unfortunately there’s nothing there. On Android, the Bluetooth toggle also doesn’t display the percentage of this Bluetooth device either. I’ve just been letting my Frame die before snapping Mister Power in to reboot it with a charge. At least there’s an LED indicator light on Mister Power to let users know that the Frame is indeed charging and when it’s completely full.
Speaking of Mister Power, there’s also a little pinhole that you need to use a sim remover tool to press and hold in order to pair the Frame with a phone (or in my original case, a tablet). It’s an extraordinarily tedious procedure that takes a bit of finicking to get the pairing to work without failing. I found that I had to actively push the pin in for the entire pairing process until after the software update, or else the NOA app will claim it failed. Long story short, the NOA app and the hardware pairing is not too user-friendly at all. Throughout my time with the Frame on my face, I’ve faced disconnection issues where the Frame and the mobile device it’s paired with refuses to connect to one another. That’s left me stranded out and about with essentially a stylish brick.
So What Can It Do?
So far all we’ve really talked about is how difficult it was to just get the Frame to function. We haven’t even gotten into discussing the capabilities of what AI on your face can ultimately do for somebody. Availability is obviously the best feature as you can’t utilize a tool if it doesn’t work. I honestly cannot say that I’ve even touched the surface of what the Brilliant Labs Frame theoretically can be used for. I just literally could not get it to consistently work.
What I personally wanted to use the Frame for was to translate menus and other items that were in foreign languages so that I could use this when traveling abroad. The dream is to look down onto a menu with foreign words, and have English translated over in my right eye. In a way, the Frame does accomplish this in a manner. In fact the first few times I asked it to translate a One Piece manga panel, I was in awe at how detailed it was. However, the more inquiries I thrusted into it, the more inconsistent the results. I began picking up things that weren’t necessarily limits, but I would classify as boundaries to operate within.
Psychologically, when you’re wearing the glasses on your face, you naturally gravitate towards the right eye as that’s where the OLED panel is spitting out text. You also tap on one side of the frame to activate it to receive a prompt. These actions make me gravitate towards my right eye as an indicator of what the Frame is pointed at. That’s something that affected my interaction with the device as the camera on the Frame is located directly on the center of the Frame. It’s right between both eyes.
I found that when asking the glasses to view and analyze an item, it would have to attempt to distinguish which item it is you’re talking about. For example, I wanted to see if it could tell me what Pokemon I was looking at from my Pokemon plush desk decor. I had three Kanto starters evenly spaced out and my eye peripheral was directed at the Bulbasaur. NOA would give me varying answers during each attempt without me readjusting my position.
Even though my right eye line was pointed straight at the Bulbasaur, the Frame camera’s field of view was able to pick up objects in a wider view. I could probably pick up the plush and hold it in front of a white wall and receive the answer I was looking for, but that’s not an ideal scenario for a pair of smart glasses that wants to lead us into an AI world. If I was in say a book shop, a museum, or a store with densely populated items, and I wanted to inquire about one particular item I couldn’t reach and relocate with my physical hands to isolate, I don’t have confidence that the Frame will give me the information I wanted to know.
To add insult to injury, NOA is an AI, but it isn’t conscious of its user’s intentions. It will always think that it’s giving users what they ask of them. If they anticipate an answer to not be satisfactory or incorrect, they’ll inquire for more information. All of these interactions eat up your limited monthly credits in the Welcome tier.
It can’t process the fact that the Bulbasaur I’m looking at isn’t a Squirtle and that I wanted to know the name of the green monster and not the blue one. I guess theoretically, one can narrow down their communication to be as descriptive as possible. Instead of asking “what pokemon is this in front of me?” it might be more accurate to ask “what is that green pokemon in front of me called?” The point is that, if I need to cherry pick my words to work with my Frame, it just isn’t smart enough to be a part of everyday human society at the moment.
Ironically, when I was filming B-roll for this this review, the Frame accurately identified the three Pokemon in succession. Just before I pressed the shutter button on my camera, I had run the same demo run and it thought the Bulbasaur was a Pikachu, and the Charmander was a shiny Pokemon (which is kind of impressive). My conclusion is that the answers NOA gives to the same questions asked in sucession are often times varying and not reliable.
I remember seeing an official Brilliant Labs video where the user asked NOA if one extra cup of coffee would be too much after consuming a certain amount already. I tried something similar by asking NOA to tell me how many calories were in an apple that I was holding. On the first attempt it told me the burger and fries I was about to eat would be a lot of calories to splurge. The second attempt properly identified the apple and designated it as under 100 calories. The third attempt to confirm the second attempt ironically incorrectly identified the apple as another fruit, a banana.
I also wanted to clarify that the Frame only displays text and there is no verbal communication between NOA and the user from the hardware you’re wearing. Inside the app that the Frame is paired to, you are able to toggle “text to speech” which will then allow the app to read what NOA is spitting out in text through your smartphone speakers. The app can also customize the style and response length in which the AI interacts with you. This is one of the cooler parts about the new generation of AI when compared to mobile assistants from the past like Google Assistant, Siri, and Alexa.
final thoughts
I’m going to finish this review by mentioning just how lightweight and normal looking the Frame is. If you’re looking at me from straight ahead, my right eye will be distorted thanks to the sliver of glass that is the OLED panel. Wearing it for long periods of time is also quite comfortable due to the plastic build. It does distort my view when the Frame isn’t displaying text because my right eye sees a prism block smack dab in the center of my field of view. I don’t know exactly how to describe it in words, but I would loosely compare it with what astigmatism would look like except to an isolated section of your peripheral.
If you pay no heed to that, it looks like a normal everyday pair of round glasses. That’s something that many other smart glasses have struggled with. The Frame doesn’t devolve anywhere with its design. It doesn’t bulk up with chipsets or battery weight across the entire structure. It really is like a normal pair of glasses that someone would wear.
Unfortunately, it also functions like a cheap pair of glasses as it is unreliable in both connectivity and AI functionality at this current time. And since the vision on the right eye is impaired due to the OLED screen, it’s not a good pair of reading glasses without AI.
This brings us full circle to when I disclaimed that Brilliant Labs called this a product for hackers. They’ve made it very clear throughout their launch of the Frame that NOA would be open source and that they were throwing hardware out there hoping for developers to make their product usable. As is now without developer growth, I just can’t see a world where the Brilliant Labs Frame is viable for normal consumers. Even as a longtime enthusiast of wearables, and as someone with plenty of experience with things like the North Focals, I found myself wanting to take off the Frame more than I wanted to have it on my face. That’s unfortunate.
We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees and support our channel by linking to Amazon.com and affiliated sites.