Competing in the post-Humane AI wearables era with Sandbar CEO Mina Fahmi

Hi, I'm Connie Loizas, and this is Alex Gove, and this is Strictly V.C. Download. Hi, happy holidays! This week's episode of Strictly V.C. Download was recorded live at a Strictly V.C. event in Palo Alto. My colleague, Marina Temkin, sits down with Mina Famy, CEO and co-founder of Sandbar, and Tony Schneider, partner at True Ventures, to discuss the future of AI wearables and what it takes to build hardware that people actually want to use. Sandbar is building stream, an AI-powered ring that acts as a personal assistant you can access anytime through natural conversation. Rather than pulling out your phone, you simply speak to your ring, whether you need to set reminders, get information or manage tasks on the go. In this conversation, they explore what makes Sandbar different from its philosophy of self-extension over replacement to its unique interaction model that keeps you in control of conversations with AI. They also discuss the competitive landscape as OpenAI reportedly works on its own device with Johnny Ive, and why Sandbar unveiled stream months before shipping in summer of 2026. Hope you enjoy it, and we will see you back here next week! The AI space has had its share of false starts. We've all been kind of following that story this year from humane $700 pin to rabbit struggles. When Sandbar unveiled its stream ring, I guess just last month, it really caught the attention of a lot of people, including Tony Schneider, of course, who met with Mina before last month. Tony has been notably skeptical of AI wearables, he's seen a lot of things, and he knows the thing or two. He was the CEO of Automatic from 2006 to 2014. He grew word for us into a top global site at True Ventures, which is a really low profile firm with a pretty great track record. He has, as I mentioned earlier, backed Fitbit, Peloton and Ring, so when he says that Sandbar has demo impressed him after countless AI devices did not, that sort of peak my curiosity. A Sandbar CEO Mina, Fami and CTO Karakong are both meta veterans who met working on neural interfaces at control labs. They spent two years building what they call a mouse for voice. Their ring now is going to capture whispered thoughts with a touch, transcribed and organized by AI. This start of his raised $13 million and is taking pre-orders for next summer. My colleague Marina Temkin is going to be talking to these guys to figure out what the recipe is to keep this company top of mind for consumers and investors alike. Thank you. Hi, everyone. It's Connie. Just introduced me. I'm Marina Temkin. Thank you, Mina. Thank you, Tony, for being here. This was such a wonderful introduction from Connie. So there is a little preview of what Mina has built and maybe we can start by Mina actually explaining what it is that stream does and also do a demo. Absolutely. We've built something that has two parts, a stream ring which I'm wearing here and stream, which we call a self extension that helps you talk through ideas and capture thoughts in the moment. Maybe I can show you a little bit about what the phone experience looks like and then it will contextualize how you use the ring. We have this little companion app but you see a feed of back and forth interactions that happen in audio. This could be about a meeting you're preparing for, it could be your driving and you're talking with your stream about an idea that you had and at any time you can say, "Hey, save that to a note and your stream will create and organize these notes for you." Normally I don't have my phone out, I have my earbuds in and the screen is turned off but I'll leave it open just so you all can hear and I'll say something like, "Hey, can you hear me?" Yeah, I hear you, you're out walking still. I'm actually at an event right now and I'm thinking a lot about what it takes for something to extend your thought rather than replace it. That's a sharp distinction. When something extends your thought, it feels like it adds reach or texture without taking over the direction. What kind of event is making you think about that? It's a, I guess like a tech in a VC event. Can you actually start a list of some taglines that might explain what you are? And as that ding happens, I feel a vibration in the ring that then creates a note filled with taglines. So this whole experience would work with your earbuds in and the phone away and is a great way to create and store ideas on the go. Every stream speaks with you in a voice based off of your own and the feeling is a lot more like talking to yourself rather than talking to an assistant or a companion, which has been the modus operandi for most human AI relationships. What was inspiration for stream? What was the moment you decided that I want to do a ring that I can talk to? Parts of it have always been really important to me. Most of my background was in neural interfaces because I always wished for technology to feel more like an extension of myself rather than a companion or an assistant. And when I learned about large language models, I thought, wow, machines can finally speak and maybe even understand. But everything that was coming out had these little quirks where you'd have to pull out your phone to type text in or you'd have to shout into your AirPods in public about really professional or really personal things, which was quite tricky. And having worked on a lot of wearables, I started to feel like maybe hardware could change the interaction model. And I reconnected with Kirak by meta control labs. And we thought, let's try a bunch of different devices. So we actually experimented a lot. And one of the early things we tried was a ring with a mic in it. And one day while I was using this prototype on the West Side Highway in New York, I started to just vent about my childhood. And I was really confused, I was thinking, I'm in public. I've never done this with an AI before, what's happening. And I began to feel like the combination of privacy by being able to whisper into the ring and the ability to hold. So I could collect my thoughts and never have the AI speak over me and the ability to tap to cut off on the syllable of what I was saying all resulted in a new conversational dynamic that was impossible with products that I had tried before. So what does stream do that can be done with existing devices? I think one big thing is that stream can act as an extension of your thoughts. And we see that manifesting in a lot of ways and a big aspect of an extension is control. With other machines that model themselves as companions or as assistants, there's this turn taking that happens where if you pause, it'll interrupt you and to interrupt it, you have to speak over it. And the nice thing about stream ring is that it keeps you always in control of the conversation. And it lets you do that in a way that is very private and respectful of people around you because you can speak really quietly and you're also providing a social signal that you're speaking. So it's not just passively recording everything around you. So if you're using your earbuds, if you're talking to earbuds, people don't see that you're talking. I think that you're actually, you're talking, right? You're just talking to yourself the same way, right? And the earbuds, there's, I guess, a few challenges. One is even just invoking the AI, often you'll have to say a wake word, which is kind of cumbersome and also the ergonomics of control are pre-uncomfortable, like fiddling with your earbuds versus the universal thinking gesture. We find changes the friction quite a lot. And also you can speak much more quietly. I'm sure everyone's experienced a moment where you're shouting to be heard into your airpods and people still can't get it, whereas with this you can be heard literally with a whisper. So there's been a plethora of AI-powered wearables introduced in recent years. Friend, limitless, osataya that is building something in a front factor of stylish jewelry. How is streamed different from these devices? I think there are a few things that intersect and I think it comes back to this idea of control or another way of describing it as of interaction. One way that Karak and I look at the arc of computing over time is that we went from a command line interface to a graphical user interface to multi-touch. And at each phase people gained new ways of expressing themselves in new contexts. And so when you want to create a new computing experience, we feel like you need a combination of hardware and software that gives people new control or a new way to express themselves or interact. And so already that prioritizes active input, like the ability to hold the ring when you want to interact versus passive recording, which doesn't really grant you a new ability. And it also means that you need a lot designed on the other side of the interface. You can't just have the device. You also need an intelligent entity that responds thoughtfully and intelligently and empathetically and does useful things for you like organizing your thoughts into notes or creating actions asynchronously. And so we've tried to treat this kind of as a holistic computing design problem, starting with a pretty narrow use case of talking about and capturing ideas that we think could expand to a more general conversational interface. Tony, investing in devices is hard. What did you see in MENA and the stream ring that convinced you that it has the right mix of design and utility to get people to use it? It is hard to invest in hardware, but we love doing it. As Connie's at air in introduction, we've had a lot of success in the category. We were the first investors in Fitbit, which arguably started the whole wearable trend. Yet we're investing in Peloton. We were investors in Ring. The video doorbell is now the largest security company in the country. Imagine hatch, just a sleep device, one in three babies born in the US, gets a hatch device. So we've seen it work. And we like it. A lot of VCs won't invest in hardware. We do. We're always looking for something great to invest in. We also believe in voice and AI devices. We were looking, we were looking at everything out there. We believe that this trend is real. People are tired of screens and phones and this is just a wonderful new way of interacting with technology. And then finally, it's just in the moment of meeting Mina, seeing the pitch, seeing the prototype. For me, it happens maybe once or twice a year. I've got a lot of pitches, but once or twice a year, you just see something I could. Wow. This really jumps off the pitch. And I honestly, I was a little jaded, a little skeptical. It's like, "Oh, not the AI device, okay, let's see." And I just walked away, like blown away. There's a true side and not everybody, but many are like, "I'm going to wear this. I love this." There's just something so easily understandable about it that really convinced us that this was the one to back. So how do you evaluate a device? Is it just something that naturally you feel this is just something as a human that I enjoy using? Is that really the final test or is there something else that you look at? That's definitely part of it. I think a big part with hardware investments is that you have to believe that the first device is going to work and knock it out of the park because you don't really get a second chance. It's not like software. We can want something and tweak it and tweak it and tweak it. Like, you have to believe that this will have an immediate impact. And believe that this is a whole category or a movement, not just some niche thing. I would also say that after that, then there's a lot of due diligence around, okay, it's a cool prototype, but can we see this get into production? Because we've made so many hardware investments, we have a whole network of people. You put me in touch with the sea like, okay, is this going to make it? Is it manufacturer or is it affordable? Does it make sense as a business? But that all comes after the initial choice of belief. For us, you just want that visceral belief and fall in love with the idea. Obviously, earlier this year, we had a big failure of humane AI-10. And I'm wondering, has that failure increased the risk for the category or the risk that? I, both. I think it's great that people try lots of devices. You have to try a lot of these ideas to see which ones fly. And I applaud any founder and entrepreneur who tries hardware because it is harder. It is more high profile. It's more visceral. People react to it a lot more. And it's harder to raise money for it. I mentioned, Peloton earlier, this is a huge success. That it's COVID-fuel peak of us worth 80 billion dollars. They were turned down by 400 VCs. It's hard. So I applaud anybody who tries, even when these devices don't work, because we have to try all these ideas. So I think what we see is there's a hunger for it. These devices come out and everybody pays attention and wants to try it. And then when they don't work, you move on to the next one. So I think it was a step along the way. It wasn't quite the right form factor, but I think it was a really good attempt. So I looked at a bunch of press around humane's failure. And a lot of people were saying as it was trying to do lots of different things at once. And a lot of the devices that are being tried right now are very focused and have a very narrow application. Would you say that stream is an example of a very narrow application because it's literally just taking notes of your thoughts, right? That's the main use case right now. I think we really wanted to focus on something quite specific that nothing else could do. And the thing that we focused on was being able to tap into these thoughts as they arise. And what we saw from computers over time was that as we went from the main frame to the personal computer to the phone, you went from being able to access maybe concepts of equations in windowless rooms to doing spreadsheets in your home to composing a song in a field. And then the missing piece was how do you tap into a thought the moment it bubbles up. And that is the very specific thing that we're starting with. But then in time, you could imagine how that can unlock all kinds of incredible things. When you can take those thoughts and move them into other applications, or you can begin to take actions in the moment that then cascade across other parts of your day. And so we've always tried to be very grounded first by the lived experience, you know, when we had our first prototypes that were half a foot long, we would strap them to our bodies and take walks with them and see how they felt. And then once we had some conviction, we would give them the friends and family to try. And always being grounded in what it is actually useful that day has helped us stay focused on the present while still trying to think about what this becomes in the future. If I could add to that, I do think it's incredibly important for devices to do one thing really well and something that people understand immediately and want to share and talk about to me, like to use a Fitbit example again, it's like when they pitched us step counting was not a thing, literally like the only way to do step counting was like a wheelie would push and would like click. And everybody's like, why would anybody do this? And they started talking about how many steps you take a day is going to be a thing and people are going to compare it and we're going to make this 10,000 steps day thing on something and you need something that's simple. And I see this with the device. One reason we're excited is I see people immediately grasp this motion like, Oh, I could just touch it and speak into it. Oh, I get it. Yeah, I want to do that. And even like at the unveiling event, Sam Bar have these clear plastic rings make just for sizing so people could decide what size to order and then people just kept wearing them and pretending to be talking into them. It's like so simple, but you need something like that and you can't be like, Oh yeah, it's going to give you directions or all these other things people think that devices are going to do that. It's going to attend different things, but not one new thing really well. Is the plan to eventually add more functionality and if so, can you give us a sneak preview what it will be? I'm thinking I don't know. Maybe there'll be biometric data and people are going to be wearing two rings. I was going to be an orange and a stream ring. I mean, there are a lot of people who love devices. I think in general, I aspire to under promise and over deliver. So I would say what we've announced so far is the very focus to use case of conversation, capturing notes and controlling your music, resuming your music, using gestures. And we're still kind of experimenting internally and with our alpha program about what we'll add next and once it feels good, we'll start sharing it. Great. Also, OpenAI is supposedly working on a device and AI device with Johnny Ive and there's a possibility, possibly, right, that they can add something like this as a feature to whatever it is they're going to unveil. What do you think is a true mode of this? Is this really the rink or is it the thought and recording the thoughts? I guess there's a question for both of you. I really believe that the philosophy you start with changes what you build. In many ways, we see ourselves as the only company really focused on self extension and human augmentation and that leads to a lot of really specific technical decisions. The fact that stream speaks back to you in your own voice. The fact that it isn't freely turn taking and that you hold when you want it to listen and you can tap to interrupt it. The fact that all of the information stream has about you as viewable and editable and something that we want to work towards is making that data freely moveable and shareable however you want so that it can feel like thoughts that are in your ownership and not walled inside of a garden. All of these come from our view of what the future should be and the role that we have to consumers as this pro-human human augmentation company. I think other businesses have their own north stars. They're trying to create autonomous entities where they're trying to serve existing business interests. I really believe those will have their place but they will take those players to a different place than what we're trying to reach. Tony, what do you think? I think this will be a big category with lots of products. I think we're already seeing it. You can buy whisperings today. When that happens, obviously if it's a good idea, there's going to be lots of people doing it. For us, it's just backing a great team that's going to execute the best in the category and so far it's just been a pleasure working with you and when we first met, it was a simple prototype and you could already see the potential but then as you start working with the team, you just notice like every step, "Oh, it's just thoughtful. There's a whole theory behind it of how all this will play out." I think it's going to take, especially with an advice you wear on you all day long, that super personal, super intimate, the execution and the details will be really important and I think this is the team that's going to pull that off and I've noticed again, I have an older prototype. It's a little more clunky but there's something that happens as you start using this. Where it feels personal very quickly and even though you said earlier you're going to focus on a narrow use case first, I immediately want to use it for everything and I think that's because it's just the product execution is so nice and I think that's what's going to set it apart but there'll be others. It sounds like you're really attracted to the form factor of a ring. That's your main kind of molt here is that it feels comfortable interacting with the ring rather than a watch or a pendant. I've tried a lot of devices and I don't wear any other wearables so I've relatively high bar for that but I feel like it's also just to want to invest in something I have to have that feeling of just falling in love with the product and I don't want to work on things that are like maybe it will work, maybe it won't or it's a good business idea but I don't like the product. I would also say one of the things that really set me apart when we first met is the demo was amazing and every time I get a pit from an entrepreneur I ask for a demo and eight out of 10 times the answer is like oh I'm so glad you want to see a demo, most investors don't even ask for it and then sometimes the day was coming, sometimes it isn't but I just feel like you have to feel the product, try the product and feel that connection to want to be part of it. I guess on the ring especially we started out building many devices in parallel and we're actually still exploring more devices but we started calling stream ring the mouse for voice because through trial and error we came to believe that it uniquely solved all of the problems that voice was facing around privacy and controllability and social signaling but I think the much more unintuitive part was that if you are to build a ring it's not enough and there are a lot of design decisions which rapidly cascade from the ring which results in stream. I may leave some of that for like when people get to have the product and experience it but maybe one way I would describe how stream has come about is with a lot of early products starting with phone based systems but even going on to future devices you saw a lot of call and response type interactions, hey what's the weather, hey order me a door dash, call me an Uber and you would have a single ask and a single reply and what we found that a ring on locks is conversation as an interface. The results in a lot of back and forth interaction something that we see with alpha testers is that 20% of their usage are these one off asks like creating a note asynchronously while 80% are iterative conversation and then there's all of the design challenges that come into creating a thing you actually want to talk to you. You'll need something that's intelligent, that's useful, that's empathetic that ideally avoids a lot of the parasocial and kind of uneasy qualities a lot of these systems have. So a lot of the attention we put into is how stream response and how it's memory behaves and how it provides information to you in a forum that's befitting an audio experience. I also understand that you can't buy it yet, right? It's coming in the summer, is that one of us, why did you decide to unveil it before the product was actually available for purchase? Well, you can preorder on sambar.com, but we'll be shipping in the summer of 26. And a large part is because we wanted to be able to build more and more in the open. We've already had dozens of people with devices in New York and San Francisco living with the product and experiencing it, but we really believe that it's going to be a dialogue that evolves in time. Practically, it's a lot easier to distribute many more test devices when people know that you exist and you're not trying to hide it. You can kind of do it openly in conversation with people. And there's also a lot of practical logistics about planning, production runs with manufacturers where understanding demand helps you do basically supply chain planning. With a lot of more mature companies, you don't need to do as much of that because you might have an existing demand model where as for a first time company, pre-orders can be a helpful way of responsible business management, basically. Wonderful. Thank you. Well, we're out of time. Thank you so much, Mina. Thank you, Tony. Thank you. We appreciate it. The Strickly V.C. download podcast is hosted by TechCrunch Editor-in-Chief Connie Loizos and me, Alex Gov of Strickly V.C. Strickly V.C. download is produced by Maggie Nye, with editing assistance by Teresa Loncon Solo and Kale. Thanks for listening. We'll see you back here next week. [MUSIC PLAYING]