You Don’t Want Smart Glasses That Record Everything You Say

Smart glasses are taking a lot of heat right now, and almost all of it is because of one thing: cameras. The idea that someone (a “glasshole,” by most people’s standards) can walk around and record others discreetly in high resolution is not an appealing one to most. Turns out, as eroded as our collective sense of privacy has become, we still don’t like when our faces wind up as fodder on some jerk’s TikTok prank.

But cameras aren’t the only way you can capture intimate pieces of one’s life without their knowledge. There’s also audio, and if you think Meta’s AI glasses are taking things too far, you’re going to want to keep reading.

---

---

Let the record show

The idea of recording everything you say all the time isn’t a new one in consumer tech. Gadgets like the AI pendant from a company called Limitless, which was eventually acquired by Meta (go figure), were designed to be always-on recorders, capturing everything you say all day long in the hopes of making you more productive, efficient, and, as a result, very creepy to be around.

While Limitless is no longer selling its AI pendant after being acquired in December of 2025, that boundary-stepping torch is being carried on by everyone’s favorite new AI gadget: smart glasses—specifically a pair made by China’s Xgimi, a company best known for selling projectors. The $399 MemoMind One are camera-less smart glasses with a monochrome display, microphones, and speakers. On paper, they’re a lot like the Even G2 made by Even Realities, which are also camera-less and focus more on productivity and notifications than entertainment, but there’s one major difference.

The MemoMind One have a feature called “Memory” that, if activated, constantly records everything you say as long as the smart glasses are on your face. Those recordings are then sent to a companion app and digested into summaries, giving you an overview of what you talked about, what you wished for, and your general goals and values.

On one hand, there’s inherent utility in something like that. You could theoretically use a gadget like the MemoMind One to remind you of things that you forgot to do, help orient your day, or even keep track of work items you discussed in a meeting. Maybe it would even have applications for people with memory problems. But with every piece of tech that captures your life, there’s a tradeoff, and that is it might be too much for most people to stomach—it certainly was for me.

A data collection bridge too far

The MemoMind One’s all-the-time recording manifests itself in a few ways on the smart glasses. The core of the feature is called “Moments” in the MemoMind companion app, which is unlocked with Memo+, a $20 per month subscription that lets you capture and store moments as well as search them from the smart glasses’ AI. Moments are where everything is recorded, and here you’ll find a summary of your day made from the audio captured by the smart glasses, which is distilled into bite-sized chunks.

In my case, Moments contained myriad things, ranging from tidbits of me talking about testing the smart glasses to my general workload for the day—mostly banal stuff. Mostly banal, but not all. In addition to cataloging my workday, the MemoMind One also tracked my general state of mind. In one Moment, I was fatigued after a long day, and that wasn’t lost on the smart glasses.

“You’ve been grinding since early morning and juggling relentless task-switching,” read one section. “The day wasn’t just about testing a device; it was about testing your own limits…”

Maybe even more concerning was the personal stuff. The MemoMind One logged my discussion about a new restaurant/cafe in my neighborhood, jotting down not just the name but also the names of two other restaurants owned by the same people—data that identifies what I like, but also most likely where I live and what I do with my free time. As harmless as that information is, reading it back in a summarized transcript felt weird at best, and that’s not even the most personal that my experience with the MemoMind One got.

While speaking with a friend of mine after work, the MemoMind cataloged bits of my conversation with her about a friend of hers who happened to be spending a lot of time with one of her exes. For obvious reasons, I won’t go into detail, but the inclusion of that discussion in my daily summary was a sobering reminder that the tradeoffs you make by wearing the MemoMind One aren’t just your own—everyone else is participating in your weird experiment, too, whether they know it or not.

And that’s just the issue; there is essentially no culpability built into the MemoMind One. While other smart glasses feature an outward-facing LED light that indicates you’re capturing video or pictures, there is no such indicator on the MemoMind One. In an attempt to offer consent to people around me, I made sure that everyone was aware I was wearing smart glasses that record everything we say, but there was obviously no real mechanism compelling me to do that other than my own moral code. As we’ve seen with other smart glasses, moral codes vary from person to person. Just because I wanted people to know that I was recording doesn’t mean other people using the MemoMind One would offer the same disclosure.

To be clear, there are privacy settings in the MemoMind app. You can set specific times for no recording, and a “Voice Print ID” feature (optional) is designed to train the smart glasses to your specific voice and minimize the chance that you’ll capture what someone else says. There are also privacy options to toggle off certain topics in the MemoMind One, which means they—according to Xgimi—won’t capture your musing about romance, “secret talk,” or legal matters, but that’s placing a lot of faith in AI to play judge. If you don’t want any memory, you should just turn the whole 24/7 recording thing off.

Still, the prospect of capturing and storing a good chunk of your day-to-day data in a cloud (Xgimi says it uses “regional servers” depending on where you are) is a bold choice, and call me old-fashioned, but it’s placing a lot of trust in one company. In some ways, the worst part about the MemoMind One isn’t its flaws; it’s that it works almost exactly as intended. When your intention is recording everything you say all day long, you have to stop and ask yourself, “Is this really worth it?”

If you can get past the creepiness…

The ability to record everything all the time is one part of the MemoMind One, but it’s not the only use. There are also quite a few of the features we’ve come to expect from non-camera smart glasses of this ilk.

You’ve got translation, which uses the smart glasses’ microphones to convert languages in real time and then displays the text on the MemoMind One screen. I tested this out with a video of someone speaking Spanish on YouTube, and it worked pretty damn well. I could see myself using this while traveling, though it doesn’t solve the problem of not being able to speak back to someone, so it would be more informational than anything. There are 26 supported languages.

There’s also closed captioning, which turns audio into text on the smart glasses display (that also works well), and recording, which can capture audio from the glasses and transcribe/summarize it in the app. The recording works in my testing, and I’m a fan of the ability to delineate speakers from one another. Similar to the Even Realities Even G2, there’s a teleprompter feature that lets you read off a script in the MemoMind One display—the smart glasses listen to your voice and then scroll through text accordingly as you read. MemoMind’s teleprompter feature is on par with the Even G2, and while I can’t see myself using this feature a lot, it does work.

Another common (and sometimes useful) feature is maps, which superimpose a small, 2D map on the glasses display for navigation. Once activated, you get a readout of turn-by-turn directions that shows how far you are from your chosen location and where to turn.

There’s also an audio input that keeps you updated as you move. Again, the result is similar to comparable glasses like the Even G2, and I could see it being useful if you really need hands-free navigation—if you’re on a bike, for example, or some other occasion where pulling out your phone would be frowned upon/dangerous. I had no issue with maps, either.

The feature (if you can call it that) that you’ll interact with most if you’re wearing these smart glasses is most likely notifications, which show up in real time as little green boxes. You can inundate yourself with notifications in the MemoMind One if you choose. To maximize my impression of notifications and make the most of my time with the smart glasses, I toggled on pretty much every possible option—email, Slack, text messages, calls, reminders, calendars, you name it. The result was… a lot of notifications.

While I’m sure there’s someone out there who would find getting notifications blasted into their eyeballs all day helpful, I am not one of them. And sure, you don’t have to toggle every notification on like I did, but even the ones that you might find useful feel kind of silly when you realize you have a phone. So many times I was just getting double-notified, since my phone would vibrate, I would look, see the notification, and then also see it in the MemoMind One display. The experience was tiring at best, and to be fair, that’s not necessarily on Xgimi—it’s a problem with all smart glasses.

Still, from a productivity standpoint—and overall execution of features—the MemoMind One is solid, with nowhere near the number of bugs I’ve witnessed in other pairs.

Good hardware, but something is missing

As I mentioned before, the MemoMind One feel very similar to the Even G2, and those similarities go all the way to the hardware level. Like the Even G2, the MemoMind One have a monochrome green micro LED display that’s binocular, meaning it’s in both eyes, and has a resolution of 640 x 350 pixels per eye with a max brightness of 2,000 nits. The result is a decently sharp and bright, if utilitarian, display that can handle daylight decently well. Menus are also fairly straightforward and easy to parse even if they lack the color of smart glasses like the Meta Ray-Ban Display.

Also, like the Even G2, the MemoMind One are very light and eyeglasses-like, which makes them comfortable for all-day wear. I weighed my pair on a scale, and they’re about 48g, making them comparable to the Even G2, which can weigh between 44g and 52g depending on your chosen lenses.

The look also makes the MemoMind One difficult to discern from regular glasses, though you can see the waveguides (the tiny screens) if you’re looking closely enough. Unfortunately, you can also see them from the inside, which creates a bit of distortion if your eyes catch them right—admittedly not my favorite experience.

While there are no cameras on the MemoMind One, there are speakers, so you can interact with a voice assistant, take calls, and listen to music or podcasts. Since I’m a big fan of using smart glasses for open audio, I was happy to see the inclusion of speakers, but unfortunately, my happiness diminished after I actually listened to audio in them.

You can listen to music and podcasts in these smart glasses, but I would not recommend it. Audio is generally tinny and poor-sounding and nowhere near the quality that I would want for listening for long periods. Say what you will about Meta’s smart glasses (and you could really say a lot), but they’re still easily the best open-ear audio experience you can have in a pair of smart glasses right now.

In that vein, call quality is decent, but I was told by the person that I called that there was an audible echo, which was less than ideal, especially if you’re comparing against Meta’s smart glasses, which are great for calling, even in loud environments.

If there’s one major difference between the MemoMind One and the Even G2, it’s inputs. The Even G2 notably come with a smart ring that can be used to control the UI of the smart glasses. The ring isn’t perfect, but it’s functional and a pretty nifty solution to a problem that smart glasses have in general. The MemoMind One, however, don’t come with any additional input option.

To control them, there’s a physical button under the right glasses arm that you can press, double-press, and press and hold to navigate menus. On top of that, there are head gesture inputs, which let you navigate by tilting your head toward an icon on a menu and then pressing the button to select. I don’t love it, to be honest—it all feels a bit clunky and slow, and, on top of all that, the voice assistant amplifies the clunkiness at times.

Like any modern pair of smart glasses, the MemoMind One have voice inputs that can be activated by saying, “Hi, Memo.” With those, you can control lots of features in the smart glasses, including notetaking, checking your calendar, referencing Moments, launching music, etc. The problem is that, in my experience, the MemoMind One are a bit hard of hearing. On lots of occasions, I had to utter the wake phrase multiple times before the smart glasses started listening. And sometimes, even when they were supposed to be listening… they weren’t.

Inputs are tough to get right for every pair of display-equipped smart glasses—it’s the reason why the Meta Ray-Ban Display come with the expensive Neural Band—but I would have liked to see better execution here, or at least more options, like a touchpad on the glasses arm. Sure, you could just pull out your phone and use the MemoMind app, but if smart glasses are ever going to feel like their own gadget, reliance on an app doesn’t feel like the way to get there.

On a brighter note, battery life is sufficient, though it will depend on what you’re doing with the MemoMind One. Xgimi says the smart glasses get about 16 hours of battery life with mixed use, which tracks with my testing. After a few days of using the MemoMind One for about half of my workday, I was at 20% battery. Even with listening to audio, that figure holds up. After an hour of listening to a podcast on the MemoMind One at 50% volume, the battery dropped from 100% to 93%, which would equate to about 14 hours of battery. You will have to use a charging dongle that connects to the back of the right glasses arm, though. Do not lose this; you won’t be able to charge your glasses without it.

A great listener in the worst way

If there’s one way to describe the MemoMind One, it would be “an escalation.” Sure, you don’t have to use these smart glasses to record everything all the time, but the fact that you can—and are being encouraged to—is something worth taking a pause over. Is this the life you really want? Would you want someone around recording things without your knowledge and consent?

Even with privacy features built in that are meant to mitigate the likelihood that the MemoMind One will capture unwanted chatter, there is no guarantee that your data will be safe—there never is on any platform—and there is also no promise that you won’t encroach on others’ privacy.

If you are the type of person who fancies turning themself into a walking, talking, human recorder, I would suggest going for a walk, touching some grass, perhaps reading a book or calling a loved one. Optimization isn’t everything, and we have to draw the line somewhere. Personally, my line ends a good deal before 24/7 surveillance machines like the MemoMind One, even if the hardware and features feel refined.