Everyday AI Podcast – An AI and ChatGPT Podcast - EP 152: Turning Snoop Dogg's Voice Into AI - The Speechify Story

Episode Date: November 27, 2023

Will the future of GenAI lead to using AI celebrity voices? How will tech like AI voice cloning come into play? Cliff Weitzman, Founder and CEO of Speechify, joins us to discuss how it's text-to-speech AI is shaping the future of GenAI and its capabilities.Newsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageJoin the discussion: Ask Cliff and Jordan questions about AI text-to-speechUpcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTimestamps:[00:01:00] Daily AI news[00:03:40] About Cliff and Speechify[00:07:20] Celebrity AI voices and Speechify[00:12:50] Safety in text-to-voice space[00:18:12] Use cases for text-to-speech[00:21:05] Audience questions[00:24:10] Future of text-to-speech[00:27:28] Is written word becoming less impactful?[00:31:30] Using text-to-speech as professionalsTopics Covered in This Episode:1. Speechify's features and applications2. Practical uses of Speechify3. Concerns and precautions regarding AI-generated voices4. Text-to-speech advancements and future prospectsKeywords:Snoop Dogg, Gwyneth Paltrow, Mr. Beast, Speechify, voice generation, deep learning, fine-tuning, celebrities, free version, premium version, reading, email, Google Docs, news, PDFs, physical documents, user affinity, AI engineering team, exceptional data, Meta, AI safety, real estate, sales, Chrome extension, web app, Mac app, contracts, terms of service, AI phone, AI smartphone, generative AI roles, learning differences, voice cloning, authenticity, identity theft, scams, reading speed, dyslexia, decoding challenge, multiplayer experience, translation, customization, books, content, audiobooks, learning, nursing programs, AISend Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info) Start Here ▶️Not sure where to start when it comes to AI? Start with our Start Here Series. You can listen to the first drop -- Episode 691 -- or get free access to our Inner Cricle community and all episodes: StartHereSeries.com Also, here's a link to the entire series on a Spotify playlist. 

This is the Everyday AI Show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life. Meet Firefly AI Assistant, now live in Adobe Firefly, the all-in-one creative AI studio. Just describe what you want to create and the assistant handles the rest, orchestrating multi-step workflows across Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome. The assistant accelerates execution. Is the future of generative AI where we have celebrities just do our talking for us?

Maybe. You know, we'll Snoop Dog be doing this intro for the everyday AI podcast someday. I don't know. But we're going to talk about what the future of text to speech and generative AI looks like with someone who's really helped building that future. So thanks for joining us. If you're new here, my name's Jordan. This is Everyday AI.

Everyday AI is a daily live stream podcast, free daily newsletter, helping everyday people like you and me, not just learn what's going on in the generative AI space, but how we can all actually use it, how we can leverage it to grow our companies and grow our careers. So very excited for today's show. But before we get into it, let's do what we always do and go over the AI news.

So speaking of text to speak, each chat GPT just rolled out chat GPT voice to all users, including free users. Yes, those of you who still aren't paying $20 a month for chat GPT, you get a new feature. So chat GPT voice, if you haven't used it yet, like I said, it was just released to all users over that long holiday weekend here in the U.S. But it essentially allows you to have a conversation with chat GPT. So think of like a Siri or like an Alexa, but something that actually gets it right on the back end. I don't know about you guys, but man, since these other Gen AI models have come out,

especially text-to-speech ones, Siri and Alexa just really don't do the job for me anymore. So if you haven't checked that out, make sure to go check it out. Next, Samsung is going all in on AI hardware. So we did report this earlier this month about the new Galaxy S-24 could be the first major market smartphone with AI baked into the hardware. So recent filings from Samsung that were just reported on today, to signal that this is more than just a rumor and it's actually happening. So Samsung recently filed for trademarks including AI phone and AI smartphone, showing that they're willing to fight for that trademark.

This is all kind of tied to their new Samsung Goss AI platform, so their own large language model that they are reportedly going to be debuting in an actual smartphone. That's right. You don't have to have a supercomputer. It is apparently coming to your actual device. So the S-24 is supposed to be the first device with this new technology, and it is slated to be released in 2024. Last but not least, companies are hiring Gen. A.I. talent at shocking numbers. So a recent study from the research firm, Lightcast, showed that there have been more than 10,000 job posting specifically for generative AI, not like one of those bullet points that are in a long list of bullet points, but specific Gen. A.I. So we'll link to this in the newsletter as we always do. But the top nine firms hiring the most generative AI talent include Chegg, Meta, Capital One, Pacific Northwest Lab, Amazon, Accenture, Nvidia, Scale AI, and Cornerstone on Demand.

So if you are looking for a new role specifically for Gen AI, some of those companies might be worth checking out. So as we always do, there's much more news in today's newsletter. So make sure if you haven't already go to your everyday AI.com, sign it for that free daily newsletter. But right now I kind of want to talk to Snoop Dog, right? I also want to talk text to speech. But it's not just me. I'm very, very excited to have on today's show. And please help me welcome. We'll bring them on here. There we go. We have Cliff Whitesman, who is the founder and CEO of Speechify Cliff. Thank you so much for joining us. Of course, Jordan. Glad to be with you. All right. So tell us maybe a little bit. If you, if you,

You haven't heard of Speechify, FYI. I'm not sure how because it is the number one text to speech app in the world. And it has men, more than 100,000 five-star reviews. And it's ranking in first place in the app store for the news and magazine category. Wow. Cliffs. So tell us a little bit about the Speechify app and what it is, what it does. Yeah.

So I'm super dyslexic. And when I was in university, I found that most audiobooks, most textbooks did not have an audio book. And so I ended up building a series of narrow applications of deep learning. This is back in 2016, 2015, that would do text and speech as a really high quality level. And I then created a series of other algorithms that did optical character ignition really well, natural language processing really well, so you could take any PDF and it would turn it into an audiobook for you. You could take a picture of a physical book and it would read that to you. You can open your emails, and it'll be that too. You can open New York Times, Medium, Wall Street Journal. It would read those

thing. So the app has now been the number one app in its category for about three and a half years on iOS for about three years on Google Chrome. It's got about 25 million people who use it. And the goal is to make sure the reading is never up there, letting for anyone, no matter what your background is. You know, I want to pull out a couple things there. One, 25 million users, right? That's insane. How did, like, what was your response, Cliff, right? Because it sounds like you kind of built this originally with yourself in mind, being dyslexic and not being able to read. I mean, what? What does it feel like now to have that thing that was built originally for you turn into

something that literally tens of millions of people around the world are using? Yeah. So there's a long history of tools that are built for people with disabilities or learning differences going mainstream. Like the best example is Edison was famously deaf predominantly and has the patent for recorded sound. And so the telephone is a great example of this. Texas speech is not a great example.

So Texas speech has been around since the 1962. It just was never good. And so I moved to the U.S. when I was 13. I learned English from Harry Potter audiobooks. And I really wanted to have a way of being able to listen more easily. And so initially I made it for me and people like me who have dyslexia, ADHD, low vision, autism, concussions, anxiety, second language learners. And you can see how those concentric circles kind of expanded out. Today, you're right.

The majority of users don't have a learning difference. It's doctors, lawyers, accountants, people in the military, executives, people in finance. I think the first like four years, it was predominantly people who had learning differences. But at a certain point around like 2021, we had this huge crossing of the Kazan because the quality of the voices started to get really good, as well as the quality of the OCR and the natural language processing. And hey, if you are late joining us this show, that's all right. It's, it's Monday morning. But we have Cliff, the founder and CEO from Speechify joining us. So make sure you get your questions in. That's a different, a little different thing that we do here at everyday AI is

You can ask questions for other people who are helping to build and shape the future of generative AI. I want to get to some of the juicy stuff, cliff, because right away, right, when you go to the speechify website, and I'm going to try to bring it up here. And if you are listening on the podcast, don't worry, you can always go and check out the show notes. You can kind of, you know, join us after the fact. But real quick, you know, Snoop talk, right? So you have all of these different voices that you can use, you know, Gwyneth Paltrow, Snoop Dog, Mr. Beas, how does this actually work, number one? And what can end users use all of this text to speech for?

Yeah. So the way that all those partnerships came together for us is typically it's people who use the product themselves who will reach out. So we had Snoop Dogg's agent freeze out and be like, hey, you know, Snoop has this amazing persona and voice. And he wants to be able to use it to help his audience consume information. And so I ended up going to his house and recorded him in his studio for a couple hours. And then the way that deep learning works in general is you take millions of hours of high quality audio.

And you build models that are super efficient. And then you can either fine tune them with either zero shot cloning or you can go more deep. And so we made the first, second, third, fourth, fifth iterations of Snoop's voice until it came out really, really good. With Gwyneth, she was a long-term investor in Speechify. Her husband, Brad, like me, is dyslexic. And over dinner, he suggested, hey, you know, it would be great if I could have my wife breed out all my scripts to me. And we were like, oh, that's really easy. And then Jimmy and I met at a party I think like three, four years ago and had to collaborating on a bunch of different projects.

And at a certain point, we decided to add his voice as well. And so what we've done is basically built a engine that is the highest quality voice generator in the world. It not only does it do it at super high quality, it does it really fast and it does it very affordably. As a result, there is a free version of speech. Anybody can use, obviously, there's a premium version as well. that gives you more, you know, quality experiences. And the way people typically will use it is the foremost highly used places are people reading their email, people reading Google Docs, people consuming the news, people reading PDFs, and then people scanning actually physical documents and listening to Speechify with

whether it be uploading Kindle books or taking pictures of actual textbooks. And, you know, for those of you listening live, I'm going to go ahead. we're going to play a preview here of, you know, just Snoop Dog. Hopefully, hopefully this works already on the live stream, but just go ahead and listen in here. Yo, it's Snoop D-O-G, and I'm an English voice from the United States. It's time for you to get your reading superpowers. Oh, I love it. Hopefully, hopefully everyone can hear that.

If not, don't worry. We're always going to link, you know, to the Speechify website so you can check that out yourselves. But so, Cliff, you actually went to Snoop Dog's house and recorded this, right? which is such a cool experience. I mean, how much validity, maybe did this not give, you know, speechify, but even just the text to speech kind of, you know, kind of genre in general, because, you know, like you said, these text to speech platforms have been around for a very long time.

Big companies have been offering them. But text to speech, like I said, it's been pretty terrible, you know, up until the last like a couple of years. I mean, what did this mean for both your company and just for the industry to have big names get behind this technology as a whole. Yeah. So Speechify had been running. We were one of the top performing advertisers in both Facebook and YouTube for a long time. We get about 100 million added persons per month. And that was mainly targeted to people who were the ideal type of users for Speatify. But really, before we started doing these types of partnerships,

speechify was a tool, like a calculator, except instead of helping you do a calculation, it helps you read. When we started in these types of partnerships, we found, first of all, people build a very strong affinity for the voices that they use inside of speech. You know, sometimes there'll be a voice that we take offline because we want to fix something about it. And the amount of complaints will get is enormous because people feel so close to the voice. And so when you have a person that you spend a lot of time with, whether it be Mr. Beast or Snoop or whomever, and you're able to carry that relationship over to academics or something

that you're curious about, that's really, really cool. The second thing for us that was very fun is we had to write the first contracts in the history of the world for licensing someone's voice, you know, for an extended period of time. where they're not actually doing the speaking themselves. And so the product started as B2C, and now we have a product that's predominantly B2B, so it's the Speechify Voiceover library. And so you're able to go to the Spegify Voiceover Studio

and you can create voice with your voice, with other people's voices, it lets you translate podcasts from one language to another. It lets you transcribe. It has an array of other features. And so the fun part here is number one. We get to build really cool technologies. There's a 30-person AI engineering team at Speechify.

My brother Tyler actually is blinded at his left He did his master's in artificial intelligence of Stanford, did his underground math there. And he built about 36 iPhone apps for the time he was, no, 42 iPhone apps at the time he was 17. And he had a cybersecurity company called Black SMS before he joined Speechify. And he ended up doing all of his master's with Speedify. And then when he joined, he took over the AI team for me. And that was one of the best days of my life because he could do a much better job than I did. But it's a lot of just honing and making sure that we're using exceptional data and that we're building a really strong team.

so we're always hiring for that team. And then the partnerships were very cool because, number one, you get to do very interesting contract formation. And number two, it's something that ends up being really, really excited for the users in the end of the day. And so whereas users previously would come purely for the function of text to speech, now they're coming because they have an existing strong relationship with a creator and they want to support the creator more.

You know, one thing that you mentioned there, Cliff, that I want to bring up is, you know, you said that you were the first company to kind of have that for, you know, the contract out there for someone's voice that they weren't actually talking, you know, they weren't actually speaking these things that you type in, obviously. But now you see other companies following suit, right? Like you were ahead of the curve. Now you have, you know, similar offerings as an example from, from meta that, you know, they just, you know, unveiled this a couple of months ago with, you know, Tom Brady and some other, you know, big, big name people. How do you view that, right? Do you view that as, oh, you know, this is a threat for speechify? Or do you view that as, hey, we know

what we're doing. We were ahead of the curve. We're a leader in this space. No, of course not. I think what MEDA is doing is phenomenal. I think that the thing to be very mindful of is you want to make sure that you protect creators and you protect people because AI safety is also extremely important. So our philosophy in general at Speechify has put the rights owners in the driver's seat. So we go and we negotiate contracts. We make sure we're doing it in partnership with people. There are some bad actors out there who will do this without, you know, an agreement and are, you know, infringing of people's IP. Like, that's very much not okay. That's number one.

Number two is, you know, there is a level of danger here. And so we're not letting anybody clone anybody's voice inside of speechify, right? If you go into the app and your provision into the beta, you can actually clone your own voice and it's super accurate. You can do the same thing for your girlfriend, for your, you know, your husband, for your children, for your grandma. And then your grandma can read your, you know, nine-year-old's book at the end of the day, which I think is amazing. but you have to have the physical device next to you. You can't just clone any voice that you get off of the internet. And so that's the thing that I really care about is I think that companies like meta adopting

this is phenomenal and more companies should do the same thing with contracts that protect the rights of the creators. But at the same time, you've got to make sure that you protect people's voices. The second thing is when you look at actors and voice actors, et cetera, so I love audiobooks. I learn English by listening to Harry Potter and I listen to about two audio books a week every single week. And so I really care that narrators, you know, get a really great experience here. And so in the same way that if you're an artist, you can use Photoshop in order to make your experience better as an artist. This is an amazing tool for narrators and actors to proliferate

their business. And like anything else with AI, it's not something that is here to replace you. It is something that is here to augment your ability to create. You know, you just talked about safety, right? Cliff. I can't even imagine, especially being, you know, one of the first major players in the space, just the amount of reports or concerns that you all have. How can you tackle this, right? Not just for, you know, kind of the, you know, the big names that you have on the platform, but, you know, what are the things that maybe keep you up at night when thinking about safety or, you know, what are maybe some recent pieces that you put into place for safety?

Because as generative AI gets better, right, and as these voices actually sound very real, there are some, you know, complications, right? So what are those things that, you know, kind of keep you up at night or the things that you're focusing on from a safety perspective? Yeah. So at Speechify, luckily, we've not had any issues whatsoever when it camps to safety. But it's because there's a lot of things that we've built in house that we still haven't even released because we're unwilling to do it in a way that is not safe.

So the first thing that we do is, like I mentioned before, you can only clone a voice if you have a phone in front of you. You can't clone a random voice from the internet. Number two is we'll only put voices on the platform where we have existing relationships and the person knows to say yes. Number three, we are internally building a classifier that if you upload a piece of audio from us or from anybody else,

we'll be able to tell you, A, this is a real voice or a synthetically generated voice, and B, here's actually where I got generated. That we think is a really good one. The things that scare me that I see online that I feel are mistakes is there's banks that still let you log in with the idea of your voice, which just is a very foolish thing to,

A, allow, B, still promote. over email that people should do. That's number one. Number two is, you know, there are these very, very scary cases of bad actors who, for example, will call someone's grandparents and say, hey, I'm calling you from prison or jail, I need bail, you know, I need help, yada, yada, with someone's actual voice. And like, this is a very high stress moment. That's really, really problematic. And so luckily, speechify is not a tool that can be used for that. However, you're in a place where with time, it's already, like, number one, it's already here. Number two, it's going to get a lot more accessible to any bad actors, you know,

over the next, like, six to 12 months. And at a certain point, obviously, like, we become more free with how we allow our software to be accessible when you can access it from other places. And so people just need to know and be aware, right? So you had this whole media frenzy in the last two elections around fake news. You'll have to have the same thing around being discerning when you see a video and you see audio clip, like you can't take things at face value. You have to confirm that it's actually legitimate. And then, you know, with family and friends, like, it's not even as far to say,

hey, just have a private catchphrase that you can exchange between the two of you to verify that you know that this person is real, this person is not real. Again, we're not at that level at Speechify so far. We're still mainly a tool that people use in order to read and write and have a better experience learning. But this is really big concerns that we have as well that we think about when we create technology. You know, Cliff, I'm wondering because you mentioned that you're still an avid reader and, you know, using Speechify for, you know, kind of the audiobooks or to read to you. What are maybe some other, maybe interesting use cases that you are still even using your own platform for that maybe aren't, you know, kind of, you know, quote unquote mainstream, you know, not just reading your emails or, you know, recapping the news for you.

But maybe what are some interesting use cases that you're using Speechify for? Yeah. So the first and most important one is they use the Mac app for Speechify a lot on the. Chrome extension for Speechify a lot. And the first thing I did is I used it to help me learn how to listen really fast. So the average American reads at 200 words per minute. I listen to most things if it's with my phone closed and off at 500 words per minute. And if the content is presented at the computer in front of me at like 700 to 800 words per minute. The reason I'm able to do this is the words are highlighted on the screen as they're being read out. And let's say I'm reading a physics

textbook. A section that I'm familiar with that covers something that I know well, I'll listen to very fast at 700 words per minute. The second I'm presented with a new equation. or a graph or something that requires a little bit more thought, I'll slow down with a keyboard shortcut, down to 300 words per minute. I'll go through that, and then I'll accelerate again. And so the ability to dynamically change the speed is the first thing that I use all the time, whether it be when I'm reading documents or emails or the news or documentation. The second thing I use a lot is actually following along with my eyes as I'm listening.

So I historically was a very slow reader, and now a very fast reader, and it's because dyslexia is not a reading disability as much as it is a decoding challenge. So if you have the word, for example, you have times. T-I-M-E, T-M-E-E-M-S-T-M-E-M-S-T-E-M-E-T-M-E. You do the phonemics breakdown. Dislexic found that very challenging. But if you see the word times in front of you and you hear it in your ears, a hundred, a thousand times, you start to be able to recognize and cite-read that word.

And so I became a really good sight-reader as a result of using Speutrify. The last thing I started to use Speechify a lot for is we're very well known for the iPhone app and Chrome extension and Mac app and Android app we have. But we actually recently launched an audiobooks product called Speechify Audio Books and it hosts 350,000 audiobooks from all the major publishers. And what we're able to do now is offer a lot of ebooks that we have partnerships with those publishers that are audiobook or AI supported. that gives the ability to normal people, especially those with learning differences, to access a world of learning that otherwise they would not have access for. So even if you look at a product like Audible, Audible has 450,000 audiobooks. But pretty soon we'll end up with a situation where we'll have tens of millions of e-books

and physical books that are accessible as high-quality audiobooks. And to be, that's something is very exciting. So much more I want to dig into. But we actually have some great questions in here from the audience. So if you are joining this live, make sure to get your questions in. If you want to know anything from Cliff about text to speech or about the speechify platform in general, but Douglas with a great one here.

So, you know, Cliff, you said that you have to actually, you know, read, you know, if you want to clone your voice as an example, right? You have to actually do it yourself. But, you know, Douglas asking here, can you add other people's voices in certain situations, like if a loved one is losing their ability to talk or maybe, you know, a spouse has recently passed or something like that? I know that there's two sides of that coin because, yes, maybe that person didn't give their consent or if someone is very sick but still alive.

I mean, what are the applications for that and, you know, pros and cons of that as well? So this is something that we're already doing for some long-term users and that we're really excited about. So one, yes, if you have a loved one who is actively losing their voice, if you just contact speechify support, either within the app or you send us an email, support at speechify.com. It's something that we're letting people do. And in that case, we're just making, you know, a case-by-case call. But one thing that we're really excited about is memorialized voices. And so we have a lot of users and we have a couple of programs that we already operate with that are for nursing.

Nurses who do hospice turns out that the things people want often when they're pressing away is to be remembered. And this is something that gives a lot of solace to folks who are at that stage of life. and especially you have family who, you know, remember them and want to be able to continue to remember them, especially when you have a large age gap between a grandparent or great-grandparents and a child who is maybe like four or five, six years old and hasn't gotten to spend, you know, as much time with that grandparent as they would have liked to. That's one thing that we're very excited to spend more time on. Another great one from Ben here.

So Ben asking, do your artists ever ask about preventing their voice from being used to read things that they wouldn't want to be used in real life. Yeah, that's a great one. Like, yeah, Ken, what if we have Snoop Dogg just promoting, you know, East Coast rappers and, you know, not the West Coast rappers? Like, how does that even work, right? Because if a big person, you know, if a big celebrity does lend their voice, there's obviously probably a lot of concerns that they have and that's probably why, you know, those contracts

are super long. But how does that work and what are the concerns there? Yeah. So there are things that we will program in that the voices cannot say. But they're not as broad as promoting East Coast versus West Coast rappers, you notice that the types of people that we tend to partner with are folks who have very, very strong followings online. And they're already known for being people who are ahead of the curb, right? Either they're very entrepreneurial or they're very creative and inventive.

And they're happy to be part of essentially this, you know, art experiment. And so typically in those situations, they're aware that that's the type of thing people are going to do. And they think it's hilarious. just introduced an entirely new way to create, bringing the power and precision of its creative suite into one conversational experience. Meet Firefly AI Assistant, now live in the Adobe Firefly app, the All In One Creative AI Studio. Powered by Adobe's Creative Agent, Firefly AI Assistant lets you start with your vision, just describe what you want, and shape the outcome as it takes form with the Assistant. The Assistant orchestrates multi-step workflows,

drawing on 60-plus pro-grade tools across Adobe Creative Cloud apps, including Photoshop, Illustrator, Premiere, Lightroom Express, and more to help bring your ideas to life. You can also get started with creative skills, a growing library of pre-built workflows for common creative tasks, like batch editing photos, creating mood boards, portrait retouching, and creating social variations. Every step the assistant takes is visible so you can refine, redirect, or take over at any time. You stay in the driver's seat as the creative director. Adobe Firefly AI assistant now in public beta. See it today at firefly.adopi.com.

Oh, man, I love it. What's one thing, Cliff, that you see the future of this space, of the text of speech? Because even myself, right, like I use text to speech a lot because I'm not a big reader, right? I like to listen. I like to learn and listen or, you know, kind of like what you said, watch the words on the screen, but listen as well. what do you see as the future of the text to text to speech space because it's obviously, you know, we kind of mentioned, you know, meta and there's obviously a lot of other

of your competitors that are out there raising a lot of money, you know, new platforms coming out. But what does the future of this space even look like? Yeah. So when I was 13, 14, I moved to the U.S. and I couldn't speak English. And when I was 18, I started university and I wasn't very good at reading. But I've consistently been the most well-red person I've ever met. It's because I listen to two audiobooks a week and I've done that for 17 years. And I listen at 3x speed and it doesn't matter whether I'm walking around or I'm in the car or I'm working out like I'm listening. And so I'm a fundamentally different person because text and speech exists.

And I ended up with this solution because I was not very good at reading. But it doesn't matter whether you're excellent at reading or not. It's just impossible to cook and read at the same time, to drive and read at the same time. And it's impossible to read as quickly as you can learn how to listen. And so you've had a huge rise of audio consumption, whether it be via YouTube or podcasts or audiobooks or double speed WhatsApp messages. And so this is a fundamental shift in information intake for a lot of people. And so essentially it's ending up making, number one, making people smarter.

Number two, leveling the playing field and making sure that people aren't painted into a stencil, especially if you have learning differences like this, like CRADHD. And so it's unlocking a lot of human potential that previously people didn't have access to unlocking it themselves. And I think it gives a lot of people the ability to dream in a way that they might. might not have been able to before, right? If you want to be a lawyer, but you're not good at reading, if you want to be, you know, in business or good at real estate or whatever it might be, like that's something that now people have access to that previously they didn't have as much access to.

That's number one. Number two is we're at the point now where people went from consuming 100 million words per month on speechify, which is a ton to a billion words per month to a six billion words per month. And so the amount of content that we're able to log is gigantic. Now, it's not just what people are electing to open on their screen. It's what they're actually electing to listen to. and we know where people click pause, et cetera. One thing I'm excited for in the long term is like TikTok,

being able to, instead of recommending to you entertainment material, material that is most relevant for your learning and for your business, given the stuff that you're consuming already. And so we've built a product right now, like I said before, that's an excellent calculator for reading. I'm excited to turn it into much more of a multiplayer mode experience where if you're inside of a team, you have a teammate who read X and a teammate who read the Y, like it'll suggest go ask Josh about this problem

because, you know, he's already consumed this piece of content. So that, to me, is very exciting. You know, quite a few people asking, so Ahmed and Carlos as well, just asking about what languages, you know, to Speechify Work in. So, yeah, what languages does it work in? So we support more than 80 languages already. So unless you speak like a very uncommon dialect, like your language is supported. Icelandic, Turkish, Chinese, Japanese, Hebrew, it doesn't matter.

Like, Portuguese, everything is on the platform. And we have a list of the dialect. that we haven't onboarded yet, and we're constantly working on adding those as well. So 95% of languages that are spoken on the world in terms of volume are already there. Yeah. Okay, good. You know, just in case I ever decide to learn more languages. I can use it.

Cliff, I have a hot take, and I want your opinion on this. And this actually hurts me as a former journalist, but I feel, especially with, you know, platforms like, you know, chat GPT or Google Bart or whatever, I feel the written, word, unfortunately, is becoming less impactful, right? And I do think that more, you know, things like this, right, like live video conversations, you know, and maybe text to speech are becoming more and more popular. Do you, do you see that happening as as well with just the written word? Do you see it losing its luster and, you know, maybe a higher priority on personalized audio or live video? To the statement that you said, not at all, but to a nuanced version of it, yes. So I'm

I like history. I was recently reading about Jefferson, and I realized Jefferson wrote the Declaration of Independence when he was 33 years old. I'm going to recently turn 30 in six months. So that to me was very inspiring. I was like those great days ahead. The Declaration of his dependence is amazing because it is so concise and so short and so to the point. And so I've always thought that great writing is great thinking. And so the written word does not lose his power in the sense that it is still the best way of expressing yourself in a manner that is concise and thoughtful. However, how that ends up being broadcast, yes, dead tree is not the best way of broadcasting your thoughts, right? Sometimes it's a video, sometimes it's a podcast, sometimes it's an audio message. Because when you broadcast, you want people to feel your tone. You want people to feel your reflection.

You want people to feel your accent, right? There's a very good saying. If you talk to a man in the language, he understands, you talk to his brain. If you talk to him in his language, you talk to his heart. So that's another thing that we're building inside of speechifies, the ability to translate all books and all content from one language to another. and doing it obviously and a voice that is close to you. And so I think that if you are a phenomenal journalist or historian or politician, the ability to consume information and then deduce what is the most valuable thing there

and then spit out the most important thoughts with your opinion on them in writing is critical. But it might be the case that that's not enough now. You need to then take to your notes and figure out how to broadcast them in a manner that is going to touch people's hearts in the most impactful way as well. Hey, great, great transition. We'll go from touching people's hearts to just hilarity. Question here from Woozy. And I'm, well, I'm going to also ask it to you, though, Cliff.

So Woozy's asking, what's the funniest voice book combo you have put together? But I'm going to also ask you, what's the voice on Speechify that you use the most? My favorite combo was when we first got the Snoop Dog voice, we played with it a lot. And Kanye West has a song. I can't remember the name of the song. But essentially, there's a second. segment there where he goes, stubri stoop-t-toop, stoop-toop-toop. And we had Snoop-Dog wrap that portion, and it was absolutely hilarious. I think we had the entire office on the floor. And so I think that

the other one was Dr. Snoop-Dog reading Dr. Seuss. It was a very, very good combo. So that's one. And then what was the second question you asked? What's the voice that you use on speechify the most? Okay, so this is a very interesting one. So Cliff Voice is also available on Speechify, and I am the third most used voice on Speechify. The number one voice, I think, is John, and the number two voice is Gretax-Palcho. One thing that we found is that if you listen with your own voice, you can actually retain the information better, because you're so used to your own voice, and it's very similar to the voice inside your head. So I use the Cliff Voice a lot, and so do a lot of my friends, interestingly enough. So I recommend, hey, try your own voice.

It's surprisingly good. Oh, I never thought about that, you know. But I also don't know, you know, having, being on a daily live stream of podcast, if I want to hear my voice any more than I already have to hear it. But maybe I'll give it a try, Cliff. People either really hate it or they love it. And for those who really hate it, I recommend listen to the voice of your spouse. Oh, okay. There we go.

All right. So we've talked about a lot here, Cliff. And maybe to wrap up, what are maybe one or two? great use cases. We've kind of, you know, talked a little bit about them throughout the course of this show. But, you know, specifically for those who are trying to grow their career, to grow their business, sometimes you don't think of text to speech, right, as that platform that can really help you grow in those areas. Maybe can you mention one or two specific use cases that can help there. The two, I would say, the first one is if you are in a career that requires

you to learn, right, whether you're a doctor or a nurse or you're in the military and you need to consume content, especially if it's like a manual or a book, throw it on Speechify, and the next time you're in a commute or you're working out or you're cooking, just click play. And make sure that the speed of the listening is equal to the speed at which your mind is working. So if you have ADHD, do not listen at 1x speed. Listen at like 2, 3x speed. That's like the most important piece of advice that I can give is use it to unlock time

that otherwise would be dead. And the second thing that's actually the most important is contracts. So if you have a real estate contract, if you have a contract with a manufacturer, if you have a contract as an actor. If you don't read your contracts, what the heck are you doing as a business professional? You always have to read your contracts. But the thing is, the thing you're looking for is the gotchas, right? So sometimes it's a hundred page contract.

All right. Throw it in there, click play at 500 words per minute and just listen. And it'll take you like an hour. But you'll notice the gotchas. And then re-listen to the portion that is the gotcha at like 300 words per minute. And it's unbelievable how often I see this use case used, especially in real estate and especially in sales. And for that, you can either use the Chrome extension or the web app for Speechify or you can use the Mac app. They're like the three best.

I love that. So there you go. Before you go clone your voice on Speechify, have Snoop Dog read their terms of service to you, right? Yeah. There you go. No one wants to read that, those boring contracts and all the terms of service. But Cliff, you bring up a good point. We should probably all be doing that.

All right, Cliff, thank you so much for your time, telling us more about Texas speech. We appreciate you coming on the show. Perfect. Have a good one. All right. And hey, as a reminder, there is a lot more. If you heard something Cliff said, make sure to go to your EverydayAI.com.

Sign it for that free daily newsletter. We're going to have a lot more in there about what we talked about and more. Thank you for joining us. And we hope to see you on another edition of Everyday AI. Thanks, y'all. Meet Firefly AI Assistant. Now live in Adobe Firefly,

the Allman One Creative AI Studio. Just describe what you want to create in your own words and the assistant handles the rest, orchestrating multi-step workflows across Adobe Creative Cloud apps, including Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome while the assistant accelerates execution. Stand control with the ability to step in and refine at any time. See it today at firefly.adobie.com.

And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going. For a little more AI magic, visit Your EverydayAI.com. and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.