Possible: R.I.P. Computer Keyboard (1964 – 2025)

0:00When it comes to AI, we spend a lot of time talking about what it takes to move models from research into reality. Enter LTX2 from Lightrix. It's an open-source audio-video foundation model built for synchronized sound and video, native 4K output, expressive motion, and precise multimodal control, all running on consumer GPUs. With over 3 million downloads on Hugging Face, LTX2 gives you full model weights, training frameworks, and evaluation tools to build real production workflows. Don't settle for a first draft. Get the complete creative engine to reach your idea's full potential.

0:40Aria Finger:Try LTX2 today at ltx.io slash model. Hi everyone, it's Rana. The future of AI is not built in isolation. It is shaped in rooms where ambitious people challenge each other. Masters of Scale Summit is back October 20th through 22nd in San Francisco, bringing together founders and innovators pushing the edge of what is possible. If you're building what comes next, you should be in this room. Apply now at mastersofscale.com slash pioneers. That's mastersofscale.com slash pioneers. Why are we still stuck with keyboards? There are 700 million people with dyslexia, and for them using technology means spelling words with a keyboard, which is the literal hardest thing for them to do.

1:34The first time people use Flow. Just give it one minute. That is all I want from you. The first dictation people do, they just speak something and they see within half a second, it's perfectly there, it got everything they wanted. When people see that, that is the biggest aha moment for them. I need to start with a confession. I am voice-pilled. Here's why that matters. Humans speak at roughly 150 words per minute, but we think at closer to 400. The average typing speed is only 40 words per minute, creating a massive IO bottleneck between thought and interface. Voice isn't just older than writing.

2:12It's been our primary communication medium for 200 ,000 years of human evolution. Next leap in human-computer interaction isn't just about bigger language models. It involves voice, thinking out loud at the speed of thought with systems that actually keep up.

2:29Aria Finger:And today's guest is one of the people making that shift real, and he has been since junior high. At 11, he built one of the world's first voice assistants three years before Siri. And when he hit his teens, he'd already built and sold his first company. And now, as founder and CEO of Whisperflow, he's behind one of the fastest growing AI products in the world, at the forefront of the post-keyboard future. I've experimented with Whisperflow, and so has my team. Here are some of the actual words from the biggest Whisperflow user on our team. Everyone I introduce to Whisperflow buys it. Take Whisperflow away from me, and I'm going to be upset.

3:07Why are you still typing?

3:08Aria Finger:So today's conversation is about that shift. Why voice and why now? And what changes when speaking becomes the primary way we interact with computers? We'll also dig into Whisperflow's origin story and the future that appears when thought and creation start to merge. Here's our conversation with Tanae Kotari. Reid and I are obsessed with Westbound and everything that you all do. So just a huge thank you for letting us be here. We have guests all the time, but only once or twice a year are we in a live podcast setting. And so we are so thrilled to be here today, especially with Taneh Kotari, who is doing an incredible job building his new company, Whisper, translating speech to incredible, beautiful, polished text.

3:54Aria Finger:So I will let Reid take it away. All right. Again, to echo what Westbound does and the community you guys are part of is extremely important. It's an honor and delight to be here. And we don't do that many of these live things. So that was part of the bow. So let's start with the kid version of Tanay. Because legend has it that you started programming at the age of 10. And secretly coding into the night when your parents thought you were asleep. So who was that kid? And what do you remember about building that first voice assistant? So the story goes way back. because when I was young, I was good at math.

4:37So what I convinced my math teacher to let me do is spend the math classes in the computer lab so I could play Pokemon. And one of those days I was sitting there, I'm fourth grade, nine years old, and there is this 12th grader sitting beside me and he's building something which looks like a new game. So I go up to him and I'm like, oh, hey, what are you doing? And he just looks at me with disgust and he's like, you wouldn't understand. Not discussed, slight over-exaggeration. But to me, it was like, no one talks to me like that. And so I go back home. I just remembered I saw it said Visual Basic on the top of the screen.

5:16And I started teaching myself on YouTube how to write code. And from that day, what started off with me just naturally being a rebel and never taking no for an answer started to become an obsession. And I started to enjoy it for just the fact that I could think about something and just build it. And then I saw the Iron Man movie and I saw Jarvis. And then I was like, that is what I want to build. And so I came back home and then I called one of my friends up. He's in the same class as me. Now we're in fifth grade. And I was like, let's build Jarvis. And he's like, that's a great idea. Let's do that.

5:55And so for the next two years, we slept alternate nights. I would wait until my parents went to bed and grab my mom's laptop because I didn't have one. And then I would call him up. We'd be on the phone call all night together, and we would just build this thing. And two years later, we launched what was then one of the world's first voice assistants. This was before Siri, before Alexa. Hit about two and a half million users in the first six months. And then Google shut us down. And it was a whole roller coaster back then, doing it hidden from my parents, but fun childhood memories. When, if at all, did they figure out what you were doing?

6:35So I had to tell my dad when I got that letter from Google and I was like terrified. It's like juvenile something, something. And so I showed it to him one morning. He was having tea. And then he was just, he had no idea I wrote code or did anything. And so then I told him parts of it. But I think the whole story I told him maybe when I went to college. I actually never slept when I said goodnight to you. Well, it's a good prep for startups. And speaking of startups, so with Whisper, are you still on the Ironman Jarvis mission? Still on that same mission. I feel a lot of what you do as adults is just what you wanted to do as a child.

7:18That makes your inner child happy when you find that right mix of your career and making the young you happy. I don't think any other feeling beats that.

7:28Aria Finger:So I have a 10-year-old, and he plays Pokemon, but I don't think he does any of the other things, but now I'm going to check on that. So you're a kid. You get the cease and desist from Google. You have 2.5 million users. How did that experience of Google shutting you down sort of shape the rise of Whisper and your interest in voice? I was initially really pissed. because I spent so much, I spent two years building this thing and one day it's just gone off of the app store. And I was mad, annoyed at them for a couple of days, but then I was like, wait, I can just build this other thing. And they just took my app down.

8:11They didn't stop me from doing anything else. And so over the next middle school, high school, seven, eight years there, probably built and shipped 50, 60 different products with friends. And these are all things that were just like, oh, I wish there was an app for that. And then we just go build it. And then for some, other people wanted it. Some other people didn't want it. And for Whisper, it all came back when my co-founder, now Sahaj, he reached out to me and he said that I'm leaving my company. I think I'm starting something new. And I was like, you know, there's this one thing I've always wanted to build.

8:49And I explained this vision to him of a personal assistant that just gets you. It doesn't make you sync time into screens, but more so just helps you be a lot more present. And if we can actually build technology that essentially just feels like a second brain, I think we can get that. And that resonated with both of us, and that was the birth of Whisper.

9:15Aria Finger:So we're always talking about how AI can actually make people feel more human. And so speaking and voice is like the oldest human technology, older than writing. It's what people have been doing forever. How does voice matter to the future of computing and technology? When I hear that, the thing I think about is that is such an obvious question where the question should really be like, why are we still stuck with keyboards? because they came as an intermediate hack in a way to help people use technology, use computers for the few decades in between where they didn't really understand language.

9:56And now that we have it, we can continue to make this interaction just a lot more seamless because the thing with keyboards is it is inaccessible to a lot of people. If you've ever seen an old person type this, this is not what their life should look like and they should not be subject to this pain and this is almost like a discrimination of some sort to being more tech savvy because it's a skill that they didn't learn growing up.

10:25Aria Finger:It's interesting. The two people in my life who use voice the most are my seven-year-old son and my 80-year-old father. So to your point, absolutely, accessibility up and down the chain. And so what do you think people misunderstand about how speaking sort of reorganizes thought and what it captures that's different than typing? So when I talk to a lot of people about why dictation, people are like, oh, you know, I type very fast. It's like, yes, you speak faster. But at the end of the day, the speed, to me, it's, yes, it's there, but that's not the biggest thing that matters. The biggest thing that matters is when you have an idea, you want your brain to go like this.

11:03like you're in the flow, your idea is coming out. But what happens when you're typing it out is, oh, should I put a comma here? How do you spell this word? Maybe I should put it in a new line. I should put the sentence elsewhere. And in the middle of what you want to just be a stream of thought, be in the flow, your brain is constantly getting distracted by the formatting and all the small nuances that come with actually writing it down. And so the biggest thing that we saw when we started people using Whisperflow, they didn't talk about how much time it saved them. They felt about how much easier life felt because this cognitive load that they had all the time, that's just like this mental process that goes on, it just went away.

11:50You just speak whatever you want and like, it's just like an assistant who like, I'll write this whole thing for you. Just like tell me your gibberish. I'll take care of it. And somebody just saying that and somebody you trust, that is the biggest value I think Whisperflow delivered to people, which I wasn't expecting, but something that we learned because that is how every single person described the product. Yeah, really amazing. And obviously, part of the thing is there's a reason why when we three are communicating, we're not sitting here typing. We're actually using our voice. And it's like, go back to that kind of naturalness about how we operate.

12:26And, of course, there's also the insanity of the QWERTY keyboard, which is, let's deliberately slow it down 30%, which comes from the mechanical typewriter. So you've seen people go from, frankly, quite skeptical to instantly obsessed with Whisper. So what do you think creates that voice-pilled kind of moment where suddenly they realize the obsolescence of the keyboard? there's a lot of reasons why people are skeptics. One, again, there's a lot of really fast typers who are like, oh, I'm like great at this keyboard thing. There are some people who've just been disappointed with voice, which is anybody who's used Siri in the last 10 years, you know what I'm talking about.

13:10And three is just people who've never seen it work. And so they're just resistant to adopt new AI and technology and all of that. And this is like, for example, my dad. The key thing that we put that does the aha moment is the first time people use Flow. What I tell anybody is just like, just give it one minute, that is all I want from you. And the first dictation people do, they just speak something and they see within half a second, it's perfectly there, it got everything they wanted, it is written exactly in the way. And small things, if you change your mind, it cleans all of that up for you.

13:51When people see that, that is the biggest aha moment for them. And then five minutes later, you know, they're telling everybody. But we realize it doesn't stop there. Once they realize that, hey, voice can work, we get them to the second aha moment is making it work for you. So what is the problem that you're just feeling in your life? You wake up every day with 50 emails in your inbox and that's like exhausting for you. Cool. Let's work on that. And when people find this first workflow that makes life really easy for them, that's their second moment. Because now they've built a habit of voice.

14:25That is when I call people being voice-filled. Because there is something in their life that they're just now doing purely with voice. Could be Slack, could be emails, could be like texting your mom, could be you are in Cursor or other tools, wipe coding. And then the third aha moment is this metric we track, which is like flow versus keyboard, how much do you use it? The keyboard just almost becomes extinct. It's less than 5 % of what people use. And that to me is what true behavior change looks like. 100%. And are there any specific features or experiences like tracking that metric that switches, that gets to that oh my gosh I have now been voice pilled.

15:13Yeah it is literally the keyboard versus flow rate like the number of characters you're typing with your keyboard versus how much you do with flow and you just see that it just goes up like this and I think across so this is something we we measured some time back, across every flow user ever, it's about 75 % is flow, 25 % is keyboard. And that has never happened in the history of computing that something has actually replaced what has been a part of our life for the last 150 years.

15:48Aria Finger:So I think, Reid, let's skip to the demo. I think this is a really good time to do that. Okay, great. And so I think Marcus is here. I mean, if you want to tee it up, what we're going to be doing right now. All right. So we've been talking about the transformative power of voice. We'll actually come back to its impact in AI and a bunch of other things as part of this. But I'm sure at least some of you are wondering, what does this actually look like in real life? So we decided to actually, in fact, do a quick demo. So today we believe voice is faster, clearer, more natural in typing. So we're going to put that to the test.

16:25And in this case, erase. Voice versus typing. So our competitors tonight are actually colleagues. Tanay and Marcus, who has proven to be an incredibly fast typist at something I can't do, 110 words per minute. Marcus, thank you for joining us. Yes, good to meet you as well.

16:43Aria Finger:So I'm going to hand each of them a piece of paper. It says the same thing. It is a chat GPT prompt. And I think Tanay wanted to make it, he was like, I got this in the bag, so he's going to drink red wine while he's doing it. He's getting a little cocky over here. Our co-worker, Sanasi, has$100 on Tanay, so if anyone else wants to take the other end of that bet, Marcus also, even though he works at Whisper, promised he wouldn't sandbag, like, going to try his best. And so we are going to watch them both on the screens. I'm going to give them the papers at the same time, one typing, one using voice, and we'll see what happens.

17:23Marcus, what's your typing speed? I type usually at 117 words per minute. Oh, so I sandbagged it in my opinion.

17:32Aria Finger:Don't open it. Don't open it. All right, I'm going to say, on your marks, get set, go. Build a Gantt style timeline product for a product launch spanning 12 weeks. Phase 1 is market research, weeks 1 through 3. Phase 2 is design and prototyping, weeks 3 through 7 with a two-week overlap. Phase 3 is development, weeks 6 through 10. Phase four is testing, refinement, weeks nine through 11. Final phase is launch, oops. Final phase is launch preparation, weeks 11 through 12. Use different colors for each phase. Show dependencies with connecting lines. I highlight the current week. I add milestone markers for stakeholder reviews in week four, eight, and 11.

18:23I didn't even know there were 11 pages.

18:30So do try this at home.

18:34Aria Finger:I mean, I want to see. Come on, ChatGPT. Make it happen. Come on. We've got to get some feedback to Sam Altman here. Oh, there you go. All right. Round of applause for both contestants.

18:53So, it's a whole stack of things. So now we've kind of introduced why voice billing is important, why it's a transformation in terms of how we're going to be engaging in compute and environment around us. But let's start with, in a whole bunch of these areas, as we get into what this means for how we work, how we live, how we navigate compute. One of the most interesting ideas is app level awareness. The idea that voice assistants can understand the context. So what does Whisper let people do that typing fundamentally cannot? Not just the speed issue, but like this turning this human messiness into clarity.

19:34It just genuinely feels like it gives you superpowers. And so imagine if there was an extremely smart person who knew everything about everything, who surprisingly happens to also be your assistant. And they're just sitting there, listening to everything that you're saying, and just doing things on your behalf. This is somebody you trust. This is somebody that learns from you whenever it makes a mistake and you correct them. And this is somebody that just has context about everything that you're doing. So it's not just ChatGPT memories that just knows all the prompts you've written in ChatGPT, but no.

20:11It knows that you were just sending an email to this person, then you're writing a message to somebody else. And the fact that Whisper is there across your devices, you use it across every single application, and it knows all of those things about you, makes it essentially this ideal assistant, very similar again to Iron Man's Jarvis, which was the dream behind it, And so the other magical thing that it does is when I'm thinking about, say, the AI adoption around the world. Us in this office, we're all Silicon Valley techies, like, great, we write our prompts, we do our things, we build our AI agents.

20:53But my dad is never going to write a system prompt. My grandfather doesn't even know what an LLM is. And so if they're using, say, an LLM to write an email, like my dad has to write hundreds of them every day, he's going to be stuck. With Whisper, what we just say is like, hey, you know what? Don't worry about prompting. There's no prompt. There's no setup. Just do what you're doing. We'll just be here. We'll look at your screen. We'll figure out this email thread. We'll write it to the right person in your style. And we'll take care of all of that. And so on the surface, it looks like a very simple product that you just press a button and it works every time.

21:27But under the hood, there needs to be so much hard work that we're doing because we're just taking all this pain away from the user. and about 20 % of our user base is what we call silvers, is like people above the age of 60. And for them, it is the most delightful AI product that they use, because it's just like one button and it just works.

21:48Aria Finger:I mean, also to your point, like if you think about accessibility globally, not everyone has a laptop, most people don't. And so it enables you actually to get so much work done or joy on your phone. And some of us, not gonna name any names, might be a little excitable and put too many exclamation points in their emails, and Whisper can help with that too, make you sound more professional. So my question is, is there like a surprising action? Like what's the most surprising action you've seen Whisper users take with your platform? So this was really surprising in the early days where, actually, you know what?