Voice Evolution: ElevenLabs’ Beta Conclusion and Tool Launch

0:00The first thing I want to talk about here is the fact that 11lab, this is the AI driven platform. It's, you know, specializing in synthetic voice technology, right? So able to generate AI voices. It has officially moved out of its beta phase and now supports over 30 different languages. This is really incredible. It's all powered by its newly developed in-house AI model. And the platform can now automatically recognize languages, including, but these aren't also the only ones, but it could do like Korean, Dutch, Vietnamese, not only can identify these languages, but also essentially promises to deliver emotionally rich speech in those languages.

0:37And if you don't know what emotionally rich speech is, you know, think of all of the terrible AI generated voices you've heard for the last 10 years. You got like Microsoft SAM default voice that's like the OG. Then of course there's Siri and even Alexa and okay, like the Google one, whatever Google assistant. These are not emotional voices. They don't have emotion to them. They sound very flat and basic. So Emotionally Rich Voices sound very human. They've got all the inflection in them. I mean, my podcast will eventually probably just get completely replaced by an AI just talking for me. Just kidding.

1:09It's actually kind of interesting at the very beginning of this. I had for about six months before launching this podcast, I had made probably 10 different podcasts, which were exclusively an AI talking and actually had some of them that really popped. Like one, it actually makes me mad. One podcast I started, I won't name it, but it's not in the AI space. It's in a different niche to promote a different business I had. But it literally, I went and looked up its analytics yesterday and it has 1500 followers on Apple music or on Apple podcasts. And the thing gets, it has around, I think like on a, on a listen basis, it was getting probably close to like, I don't know, a 20th.

1:54So like 20 % of what this podcast gets, this one's absolutely blown up. This one went crazy, but it's getting 20 % just from AI. And that was like after a number of months. So, and like really not that high quality podcast, I'm actually going to have the whole thing completely converted to actually having a real human run it. But I guess I tell you that because AI voices are really cool and a really good way to do a lot of stuff. But I do think there's a lot of value in human voices. So when I started this podcast, I was like, I'm not going to use an AI. I'll use a human voice. I'm hoping you guys appreciate my antidotes and like commentary experience in the industry, blah, blah, blah, all that kind of stuff versus just an AI that, yeah, ripped something from Chad CPT or something to read to you.

2:35In any case, I think what really sets, getting back on topic, I think what kind of sets 11 Labs apart is its voice cloning tool. If you haven't used it, it's incredible. I've used it, have it set up for a bunch of voices for myself. Again, don't worry. I'm a human. I haven't replaced myself yet. But in any case, the feature really allows users to speak in over 30 different languages without the need to type text first. So what's really cool about this technology, in my opinion, is the fact that if I cloned my voice, I could, of course, be talking to you in English, but I could use an English transcript or English audio files to clone my voice and have it speak in a ton of different languages, but it still sounds like me.

3:14So personally, I speak English and French. I lived in France for two years. But what's interesting is like, I wonder what it would sound like if I trained it with my English voice and had it speak French. I bet it would have a better accent because obviously I have an English accent. I do my best, but not perfect there for sure. But I wonder what it would be like having to speak French for me. I think that'd be very, very interesting. So in any case, the CEO and co-founder of 11 labs expressed the company's vision and said, quote, 11 labs was started with the dream of making all content universally accessible in any language and in any voice.

3:51So with this release, we're one step closer to making this dream a reality and making human quality AI voices available in every dialect. Our text to speech generation tool helps level the playing field and bring top quality spoken audio capabilities to all the creators out there. So this was, Eleven Labs was founded by Steinswiski, a former Palantir employee, and his childhood friend, Piotr Davoski, who is an ex-Google staff member. And Eleven Labs was inspired by the really subpar dubbing of American movies, which both of them encountered growing up in Poland. Shout out to the people of Poland.

4:31And I'm sorry, you had to listen to our terribly, you had to watch our terribly dubbed movies. I got a lot of great friends in Poland. But the platform really quickly gained traction since its beta launch in late January. Thanks in part, I think, to kind of the quality of its generative voices. And also it had a pretty generous free tier that I think helped it grow a lot. So in any case, this has not always been a smooth ride for the company. I think the platform actually faced a lot of controversy when it was exploited to share, you know quote unquote hate messages on forums like 4chan and essentially create problematic audio samples so in response to this 11 labs introduced new safeguards including limiting voice cloning to paid accounts and launching a new ai detection tool it's kind of interesting a lot of times when these companies get controversy you know mid-journey did the exact same thing it's like the first thing they do they're like okay it's not free for everyone anymore and they so the secret there in case anyone doesn't know because i mentioned this before in the podcast but the reason why um making it paid and not having a free tier anymore for things like voice cloning is a good like safety feature is because it's not just because like oh it's like paying people will still break the rules and do things they're not supposed to but the benefit there is for the free tier you could be anonymous and so if you pay you have a credit card on file and unless you're committing credit card fraud which is a whole nother problem um then if you're breaking the rules they know literally who you are they can ban you they can ban your name from trying to do things in the future.

5:55And if you were making it say something super bad or illegal or blah, blah, blah, whatever, like you're a lot more, they pretty much have your identity if you are a paying user because they got your credit card. So that's kind of the concept there. In any case, the platform's really rapid expansion, I think has also ignited a lot of concerns about the future of the voice acting industry. That's what, you know, right now we have the Writers Guild is on strike because Chad Chubut makes them, you know, worried they're all gonna lose their jobs. Well, guess what? voice actors are definitely at big risk right here with Eleven Labs.

6:27It's an incredible technology. It does a great job. So that's that. Love it or hate it, that's what it is. I think voice actors are kind of increasingly finding themselves being asked to relinquish the rights to their voices, which really would enable clients to create an AI-generated version of their voice. And companies like Activision Blizzard are also exploring voice cloning tools, which has raised a lot of questions about kind of the long-term impact on the industry. But despite all of these controversies 11 labs has shown no signs of slowing down this is actually i will be i'll i'll give a big um i'll give a big congratulations to 11 labs a big compliment in a sense that i remember i mentioned all of those ai podcasts i've used i actually did not use 11 labs to generate them i use well said labs um maybe their their pricing structure was better or something but the reason i used 11 labs is because uh or i used well said labs is because they had a voice that sounded just like it was like uncanny how human it sounded.

7:20They have one voice in particular. It was really good. 11 labs. I also found did, but I think maybe the price scene was more expensive or something. But in any case, it's interesting because I, if you've looked at my podcast, you've seen, I've covered 11 labs like a dozen times on this podcast. And I have yet to cover well said labs. Like I'll mention it now, but I'm not seeing it like in any news articles or in any kind of top of the news of new features. And I'm sure they're rolling stuff out, but they're just not like doing some of the big pushes that 11 labs and others are doing. They're not raising the big money that a lot of these others are, are raising.

7:50And so it is really interesting. I will give a big shout out to 11 labs. Like they really are pushing the envelope and they're really like pushing hard as far as innovation in the sound space goes. And, uh, yeah, I think that they're doing a good job. Like, honestly, I think 11 labs is a good job, a good product. The voice cloning is super cool. So big, big shout out to 11 labs for all of that. I think that, um, it has collaborations with publishers like Storytel and some media platforms like The Soul Publishing for audiobooks and also radio content. And then they also have over a million registered users across a bunch of different sectors, including creative entertainment and publishing.

8:27And all of them have collectively generated a decade's worth of audio content, which is pretty impressive. So I think recently they secured$19 million of investment from Andresen Horwitz, right, A16Z, and also DeepMind co-founder Mustafa Suleiman, and they did this at an evaluation of around$99 million. So Eleven Labs has very ambitious plans for the future. I think it aims to really extend its technology to voice dubbing, following the path of startups like Papercup and DeepDub. And I think the ultimate goal for them is to build kind of a foundation to be able to transfer emotions and intonation from one language to another.

9:02And then further down the road, I think Eleven Labs intends to introduce a feature allowing users to essentially share voices on its platform. Although the specifics remain kind of unclear at this stage, but you can imagine how popular that would be if there was a marketplace where like as a voice artist, you could just clone your voice, put it on there and allow people to like rent your voice essentially, right? Pay you for the usage. You get a royalty. I think this is, that's an incredible platform and that will be super popular in the future and save voice artists a lot of time, right? They have a great voice.

9:31They get paid for it and as they should, but you don't have to sit there and manually record everything. Like anyone complaining about that has not spent hours and hours recording their own voice like I have for this podcast only because I love you guys though. Um, and it is a lot of work though. And it's a grind. And honestly, some days after recording a bunch of podcast episodes, like my voice is sore and it would be amazing if there was a tool that could overcome that for, you know, commercial uses and other things like that. So it's going to be cool to see what happens, but, um, this is definitely an awesome company that I'm a big fan of.