Loading summary
Transcript1 lines
- [00:00]
A
Finally, the moment I have been waiting for for like freaking 4 months. OpenAI has stopped dragging its feet and is starting to roll out their advanced voice mode. This is something that they showed off in their spring update, like four months ago. There's a few things that are still missing. One of them is the video that takes over your whole phone. There's a bunch of other things I'm gonna be telling you who this update is coming to, what, what's going on, what is enabled, and everything all about it. But before we get into that, I wanted to say that today's episode is sponsored by my very own AI Hustle school community. So if you've ever wondered like, or if you've ever wanted to use AI for a side hustle to make money or in your career, if you want to level up your career or with your business, if you're like, gee, I want to know how to use AI to make more money, scale my business and grow faster, you need to join our school community. It is going to be a hundred dollars a month, but we have it at the, we have it at a fraction of that right now. I think it's like 20 bucks a month. And if you lock in the price now, we're not going to raise it on you. It's an incredible community and we cover everything that you need to know to use AI to scale your business or to make money. So Link is in the show. Notes to the school community. Would love to have you in it. Let's get into what OpenAI is doing. So this is actually kind of crazy. The, the thing that I think is amazing about this, that I was the most impressed with when they demoed this entire feature set from OpenAI is the fact that right now we're all used to AI voices. This isn't like a new concept. They already have AI voices on chat GPT that you can chat with. There's like two or three AI voices and those are kind of the traditional AI voices that you see everywhere. The things like 11 Labs or well said labs, both of these I've used to run entire podcasts. Not, not my own podcast. Don't worry, this is not a clone today, but maybe someday, who knows, right? But I've used like those other kind of AI voices to scale up entire podcast, like over a hundred thousand listeners before. So, like they're good enough that people like them and people, all sorts of corporate people use them. This isn't what OpenAI is doing. They have a whole new thing here that to me, the most impressive Part was the fact that essentially the voice is very dynamic voice. So it's not just trained off of a bunch of audio files that are saying something in one specific way. For example, if you took every podcast I ever recorded, threw it into a regular voice model, it's going to sound pretty like if. And if I just typed in the words, it would say something pretty similar to this. Minus, like my stuttering or me pausing to think or me saying or. But. Or like, I don't know, stuff like that. Right. So you wouldn't. It wouldn't be quite as natural. So what OpenAI has done though, is they've trained a model that's super dynamic, meaning that you can tell it to, you know, give it a script and say, hey, like, say this with the voice and it will say it. Then you can say, okay. Say it like you're running up a hill and you're like, out of breath. And it's like, okay, I'm coming. Like, there's no way you can train a normal AI model to do that. Well, the ones of the past, they were not able to do this, right. They could do all sorts of crazy things. I've seen some pretty crazy demos. People say, like, pretend that you're on like a peloton and you're like the run the, you know, the bike instructor, and you're like telling everyone to do everything and you're out of breath. And it does it. Um, it says, you could say like, say X, Y and Z. It says, it's like, okay, now sing it. And then it sings. It's like, sing it happier. Sing it sarcastic, like, more sarcastically, like, now be angry when you say it like. So it has all of the intonation, all of the intonation, all of the inflection that's all baked into the AI model. So it's, it's phenomenal. It will blow your mind if you listen to it, if you try it. And finally, OpenAI is rolling this out to everyone to see. Now, when I say to everyone to see, obviously that's not actually to everyone. This, this is going to be just to the pain users. I was super stoked because I pay for ChatGPT, like, probably a lot of you. And I hopped on my phone and I looked and all I got was this stupid pop up. So it's the. I'll. I'll give, I'll give you the lowdown on this pretty much. It says advanced voice mode is on its way. A roll of advanced voice mode has started and we're slowly rolling Users in the alpha to enrolling users in the alpha to ensure the quality of the experience. All plus users will have access by the end of fall. I'm gonna let you know as soon as you're in. There's a little okay bubble. Right. So moral of the story is by the end of fall, which what's this gotta be like probably the end of October, right. I'm hoping everyone should get it, but they're rolling out, it's coming. So rest assured the wait is over four freaking months and we're finally gonna be getting this. So I personally am pretty excited about this. They had a tweet about this which is kind of funny. I'll tell you why it's funny, because it's. And this is from the official OpenAI account. They said advanced voices rolling out to All plus and teams users in the Chat GPT app. Over the course of the week, while you've been patiently waiting, we've added custom instructions, memory, five new voices and improved accents. So they added a bunch of new stuff which personally I would have rather they just release it when they announced it and added these things as they went. I think it would have gotten more hype and been more exciting. But whatever they then they said in the tweet, it can also say sorry I'm late in over 50 languages, which I think is fitting. And I think that they should apologize for making this thing four months late. But whatever, that's just me. So what are the new voices that this thing's getting and who else is kind of doing this? There's five new voices which I'll tell you the names of them but essentially this is useless because it doesn't explain much until you can actually hear them. But they have Arbor, Maple, Soul, Spruce, Veil. Right. They're doing all like the nature name things and they already have Breeze, Juniper Cove and Ember, which yeah, whatever. Okay, so that's cool. The thing that's funny though, so they now have like nine names. What's interesting is Google Gemini Live also does a lot of voices. Now it doesn't quite do the same dynamics thing, but Google Gemini Live has a little bit more than that. So still sort of beating them on the number. But I don't think the numbers really matters that much because if you go to something like well said labs or Synthesia, they got tons and tons of voices or like 11 labs, tons and tons of voices, but doesn't necessarily mean they're better. I don't think anyone is doing this kind of dynamic voice changing that they're doing, which is absolutely insane. So it's funny because they're. They're doing. All of these things are named after nature. And I think it's kind of like, you know, a plug at like, look, we're, like, making it more natural, which it is. It's. It's impressive. The one funny thing, the one controversy in all of this is that there's a voice that's missing that they announced in the spring update. Oh, what happened? And that is the voice of Sky. So sky, for those that don't know, everyone, everyone that tried the. The voice, sky, it sounded exactly like the AI system from the movie her, if you've seen it, which essentially is just Scarlett Johansson's voice. So there's a lot of controversy that goes into this. And it's. It's just a crazy coincidence that it sounds really similar to her. But essentially what happened was, I think like Sam Altman or OpenAI texted Scarlett Johansson and was like, hey, can we, like, use your voice and train a thing on it? And she didn't, like, give them permission and they wanted to do it because obviously the movie her with the AI voice assistant, they're like, oh, this will be, like, cool. This will be kind of meta. Right? She didn't give her permission. And then they released their model and showed their update with that voice. Then it got tons of controversy. I think she threatened to sue them or sued them or something over it, and then they deleted it. And we're like, no, it's a crazy coincidence that sort of sounded like you, but we're just deleting it for no apparent reason, you know? Right. So obviously they probably. Probably got caught up in that one a little bit. But the voice is gone. Many people are sad about it. We will continue to move on in any case. This is pretty awesome. I'm super excited for everything that's going to be rolling out. I will keep you up to date. Now, the one thing that I will say, if you are waiting and excited for this, you are not going to get it in a few different regions, or at least it's not yet available. That is the eu, the uk, Switzerland, Iceland, Norway and Lichtenstein. I am so sorry, everyone from Liechtenstein. It's not available yet, but it's coming soon. I think the bigger thing here, though, is that it's not available in the eu. I'm really hoping it rolls out, like, without any issues in the eu. I'll keep you up to date. Have a fantastic rest of your day.