AI transcript
I think the big a-ha moment for a lot of people is gonna be that probably in the next six months to a year that the main way
People are using AI is voice
Hey, welcome to the next wave podcast my name is Matt Wolf
I’m here with my co-host Nathan Land and we are your chief AI officer
It is our goal with this podcast to keep you looped in on all the latest AI news the coolest AI tools and
Set you and your business up to be ready for this next wave of AI that’s coming and today
We’re going to talk about what we think is actually the next wave of AI your voice in this episode
We’re gonna talk about why we think voice is the next big thing
We’re gonna talk about the tools that are available so that you can use voice
We’re gonna talk about the scary risks and why this could be fairly damaging to the world if things go in the wrong direction
we’re also gonna talk about how AI is taking over the music world and
What potential options are out there to solve some of these problems?
so a really really fascinating episode lots of cool stuff we’re gonna talk about and
Excited to share with you. So let’s go ahead and just dig in
Yeah, so today when everyone thinks about AI most people are thinking about text using text to talk to AI
You know like chat GBT Claude
but I think the big a-ha moment for a lot of people is gonna be that probably in the next six months to a year the
Main way people are gonna interact. I believe is gonna be with voice, you know
And this company Hume AI just raised 50 million dollars in their series B
And they look they just launched a product called Evie, which is a empathetic voice interface and you know
I tried it yesterday with my son and
And he was blown away, you know, he’s ten years old
He tried it out and he was like what is that and he saw like the little line on it where it’s showing the emotions
You know and he started and he started talking to it and he was just like freaking out because it was
Detecting the emotions in his voice and then responding accordingly
When all your marketing team does is put out fires they burn out fast
Sifting through leads creating content for infinite channels endlessly searching for disparate performance KPIs
It all takes a toll but with HubSpot you can stop team burnout in its tracks
Plus your team can achieve their best results without breaking a sweat with HubSpot’s collection of AI tools
Breeze you can pinpoint the best leads possible capture prospects attention with click-worthy content and
Access all your company’s data in one place no sifting through tabs necessary
It’s all waiting for your team in HubSpot
Keep your marketers cool and make your campaign results hotter than ever visit hubspot.com slash marketers to learn more
No, it’s really interesting. I played around with you
I actually made a YouTube video where it was like the AI news recap
But in one piece of it, I played around with Hume for a minute
And it was kind of funny because I was like hey Hume, how you doing today or something like that?
And it was like I I detect that you’re angry. Why are you angry and I’m like I’m not angry
I think I was probably like when I make videos and when we record podcasts like this
I tend to ramp up my energy and you know
I’m probably a little bit more excited than if I was you know
Just talking to you in real life face-to-face and I have a feeling Hume sort of picked up on that sort of extra
Excited energy news like what is wrong with you? Yeah. Yeah. Yeah. Yeah. Yeah, but anyway, you know
Back to the bigger point here is that I do feel like we’re entering this world where right now
Most of the prompting we do is with text prompting you go to chat gpt
You go to Claude you go to one of these platforms right and you type in what you want it to respond to
But I feel like over the last decade or so the tech world has kind of been training us that know your voice should be
Mechanism for interacting with the computer right with things like Siri and what’s Google’s version Google’s assistant version
Is it not Google home? Yeah, but anyway the point being like yeah, we have all of these smart devices
We’ve got our phones. We’ve got our Alexa’s. We’ve got stuff like that
That is I think where things are going right you’re going to be actually speaking to these things instead of typing prompts
and with things like Hume, we’re gonna actually get even more context for it to analyze when it gives us feedback, right?
The rumor has it that at WWDC this year Apple’s big developer conference that they do every year
That’s kind of where they make the big announcements
It was WWDC last year where they announced Apple Vision Pro this year
The big expectation is going to be a lot of AI right and I think a lot of people are speculating that Siri is going to get AI
Into it what they’re gonna use for the AI. We don’t really know yet
You know, we we do know that there’s been rumors going around that Apple and Google might partner and that Apple might use Gemini
1.5 for Siri behind the scenes, but then also
Apple just released a research paper the week that we’re recording this apple just released a research paper called realm
And it’s basically their large language model that they designed for a mobile phone. It is actually very good at reading the context of what’s on the screen
it can actually view a mobile phone screen and
Use that context to sort of inform the large language model before it responds to you
So a lot of people are thinking well, maybe that’s what’s gonna be in Siri, right?
So maybe you’re on an app. You’re using the app. You might be able to go Siri. Hey, this app isn’t working
Can you help troubleshoot it for me? Siri will actually be able to see your iPhone screen see what’s going on and
Possibly help you through whatever issue you’re running into if they use this realm language model that Apple just put out the research for
But where they actually go with it. I don’t think anybody’s gonna actually know until WWDC, which is I believe in June of this year
I think it’s crazy that like, you know, Alexa and Siri both came out like 10 or 11 years ago right around the time when the movie
Her came out, you know, everyone thought this technology was gonna, you know, improve every year. It was gonna be amazing
And so a lot of people I think they think oh this stuff is not gonna get better because look look it took 10 years
It’s still almost the same product
But they don’t realize like when like GPT-5 comes out if you built something like Alexa now with like GPT-5 under the hood
How dramatically different of an experience that’s that’s gonna be and and I’m sure Alexa will be upgrading soon too because you know
Amazon just invested. What was it like four billion dollars and then something like that. Yeah
Well, they initially invested like 1.75 billion or something and then when yeah all this data came out that
Opus is actually beating the latest version of GPT-4 in all of the benchmarks
Yeah, Amazon came back to Anthropic and said hey, maybe we’ll put even more money into you guys
And then I think they invested like another two and a half billion or something like that, right?
So we’re gonna have Jarvis in our house pretty soon if you want to yeah
And if you think of that from like a business context, so you know as a CEO or a manager like how amazing is that gonna be that instead of
Having to type everything to your AI you can be like really quickly like hey
You know this is going on or hey, can you check this report for me or can you check?
You know what do I have an email from so-and-so and can you respond to that doing all of that by voice like how much more
Efficient you’ll be as an executive using these kind of tools. Yeah
While we’re on the topic of voice to open AI just put out some new research as well, right?
Open AI just put out an article on their blog saying that we’ve got this really really good voice model where with I don’t remember what it
Was exactly but like 15 seconds of training data. We can replicate your voice and then you can type in whatever you want
It’ll say that thing. Yeah, but then open AI in that same article wet, but it’s too powerful guys. It’s too strong
It’s too scary. We can’t let you use that yet, which to me is just wild because we’ve already got it, right?
We’ve got 11 labs 11 labs already do exactly that right we’ve had like uber doc
There’s been there’s been a whole bunch of models out there already that do exactly that an open AI kind of coming out and saying like
We do it too, but we’re not giving you our version because it’s too good kind of weird
So when I saw Evie from Hume, I was like, okay, that’s probably like the best I had seen so far like on the emotional side
That’s it’s amazing to texture emotions, but you could still kind of tell it’s not a human and with 11 labs pretty good
But you can still kind of tell it’s not a human. I know that the demo I saw I believe it’s from open AI
That was the most realistic one I had heard. I was like, okay, that’s like 99.5 percent there
Yeah, like if I didn’t know this technology would existed. I would not know that was AI
I would think that was a real person talking. Yeah. Yeah, the other ones are like 95 percent there
But when you get to like 99 like people just they don’t know that it’s an AI and that’s where it’s you know
I can get that. Yeah, that could be kind of scary that you could just be pretending to be anyone
Yeah, no one would know well totally
But I also think that you know people like you and I who are immersed in it on a daily basis are probably a little more
Adepted actually spotting AI right like I feel like I’m really good at when it comes to AI images and going I can tell like
Instantly that it’s AI now because I’ve seen so many AI images
I feel like I’m getting that way with yeah voice now to where I can pretty quickly grasp
Okay, that was done with AI
But I mean how many people have actually been fooled already by AI voice using things like 11 labs
I mean that that that kind of sort of brings us into the I guess kind of the next thing that I want to talk about is like a
Scary part about all of this AI voice technology that’s that’s coming out right now is you know
there’s there’s been stories of a
Scammer calling up somebody’s parents and saying hey
We kidnapped your daughter and then they would use a voice sample of that person
to convince the person on the other side of the phone like oh they really have my daughter and
Collect ransom money right. There’s also that what was it the 25 million dollar scam that happened over zoom
Like remember when we talked about that for a little bit
Yeah, yeah, like so I mean that yeah stuff’s happening right now
Yeah, and I think that was here in Japan. I believe and so yeah somebody
impersonated an executive or I think they print impersonated with the CFO or was like a top executive at a big company and
Called people up and people didn’t know that it wasn’t a real person
And they end up wiring. What was it 25 million or something or something like that?
It was a lot of money that they just wired off to some scammer. Yeah
Yeah, the tech to do the trickery is already out there
Yeah, but like you said like yeah open AI is the best one to come along so far
Yeah, but it’s already out there like people can already do this right now with it with this kind of technology
So there’s already laws that this this kind of stuff is illegal like so I guess the challenge is the scale will probably greater
Because like it’ll be easier to scam people now. Yeah, right. So and that is probably an area where you’ll need AI to help like
You know find scammers and things like that, right?
Which you know hopefully doesn’t go into like big brother territory where we’re using AI like monitor everyone
But you know possibly that’s it is slightly going towards that direction like yeah
You need AI to like figure out like so-and-so scamming people and then yeah look at the data and then go get the person
Yeah, I’ve got somebody that like is anti regulation. I actually think regulation to some degree is of a good thing
In the sense that I feel like regulation gives companies bumpers to stay within I think that could be a good thing
Right, I think right now. There’s a lot of AI companies out there. They’re developing and they’re going well
We don’t know what’s gonna happen with regulation
So let’s just keep on pushing the limits
But I do feel like regulation to some degree gives them some bumpers to stay within so they know they’re not overstepping their bounds
However, when it comes to AI, I don’t feel like regulation works at all
Because bad actors are gonna be bad actors, right? So we can go out there and say like hey
We’re gonna regulate you’re not allowed to clone people’s voices with AI and do people
Well, you already can’t do that, right? Whether it’s law or not, you know, it’s already highly unethical people already see that as a negative thing
I’m pretty sure it’s already a law that you can’t do that anyway
But yeah, yeah, the technology’s out there. It’s out there in open source form. It’s out there in closed source form
It’s out there people are gonna be able to do it. So how do you like what is regulation going to accomplish in that sense?
I don’t really understand the regulation argument here
You know, most of the politicians have very little like real-life business experience
And so like when they think about regulation a lot of times it’s just like, okay, is the public scared of this technology?
Well, okay, then I’m gonna regulate it. It’s like it’s actually good for your country. Is it good for like business?
Is it good, you know, are you just literally responding to some polls? Is that all that is and you know without going too deep down like the sort of
Political rabbit hole there’s lobbyists, right? And a lot of the government is run by the companies who are paying
For those people to be where they are in the government, right?
So what I think what the big fear is and I know this is something whether you love them or hate them that Gary Marcus has talked about a
Little bit in the past. Yeah, is that what happens when we get too much regulation with this is what it’s gonna end up
Doing is concentrating the power into a few small companies, right?
What’s gonna end up happening is the the companies like open AI and Microsoft and maybe Google and some of these big corporations that are
Pushing for the regulations may end up driving the regulations in ways that really favor their company
But really do not favor the little guys, you know, that’s kind of like the biggest divide that’s happened in Silicon Valley recently
Is like you got the divide between like the Sam Altman, you know side
Which is also kind of aligned with the YC and then the Mark Andreessen VC side where they’re like, yeah
This is basically regulatory capture here. Like they’re like trying to go out there and like say yeah
What we’re building is very dangerous. So you should please regulate. Yeah, that’s an odd thing to ask for like why are they doing that?
Please regulate us and we’ll be on the board to decide what those regulations are
Yeah, yeah, yeah regulate us in this exact way that we want that no startup can afford
Yeah, and oh by the way, open source is very dangerous and so that’s kind of been like the undertones of it all too
It’s like when they’re talking about regulation. It’s like open sourcing of AI is very dangerous
That’s almost always the undertone. It’s like well, if you don’t have open source AI
Then yeah, you will end up with like one or two companies that controls all the technology
So so that’s why I’m like very hesitant of like like I’m not gonna say there’s no regulation
They did like maybe there are there’s regulation needed for like, you know deep fakes or you know AI voice at some point
But like just speeding into it and trying to regulate everything kind of like what’s going on and you’re a little bit right now
I’m like, that’s not the right approach
Hey, you know a lot of a lot of where my heads at to kind of comes from like the crypto space, right?
Where there’s been looming regulations forever, but the a lot of the regulations never end up happening
So a lot of the companies that are trying to build or like am I gonna build something that’s gonna end up getting regulated out of existence?
So like a lot of people and companies in that crypto space are like just tell us the damn regulations
So we know what bumpers to stick with it and we’re not in this limbo
But when we’re like talking about AI, I feel like it’s a different story because it’s like you do have the open source
You do have the closed source
You do have the companies that you know have a better foothold within
Governments than the little guys you do have all these other nuances
I think really muddy the waters and also I feel like regulation is just gonna make it easier for the bad actors to be the bad actors
While the the people they’re trying to do right lose abilities essentially right along with all the negatives that we just talked about
There are still some positives of this this voice AI that’s coming out
I think there are some use cases that I think could be very valuable
I know for me as a content creator using it for like dubbing is really really helpful like if I misspeak in one of my videos
I don’t have to go back in record and like overdub something
I can open up a tool like descriptor 11 labs type in the words that I meant to say and then just use that and dub it into my
Video so I’ve actually used some of these tools to sort of fix a misspeak in some of my videos
So that’s like that’s one really good use case for some of this AI voice
We’ll be right back but first I want to tell you about another great podcast you’re gonna want to listen to it’s called science of scaling
Hosted by mark robert’s and it’s brought to you by the hub spot podcast network the audio
Destination for business professionals each week host mark robert’s
Founding chief revenue officer at hub spot senior lecturer at Harvard Business School and co-founder of stage two capital
Sits down with the most successful sales leaders in tech to learn the secrets strategies and tactics to scaling your company’s growth
He recently did a great episode called how do you solve for a siloed marketing and sales
And I personally learned a lot from it. You’re gonna want to check out the podcast
Listen to science of scaling wherever you get your podcasts
Yeah, and human they’re kind of pitching that this is gonna be used for therapy and things like that right which is
Exciting you know, it’s kind of you know gonna be weird to talk to a robot
And that’s how you’re getting your therapy, but I mean who knows like when it gets good enough
You know, I guess that’ll be a thing it does kind of make me think of the movie her
Yeah, right, you know, we’re you’ve got this guy who I think you haven’t seen it
And I feel like you should you should take my AI fan card from me. Yeah, you seem ex Machina. I have seen ex Machina. Yes
Okay, okay. Okay. You’re okay. You’re like you’re halfway there, you know
So with her you’ve got you know, Joaquin Phoenix playing this guy who’s like got divorced. He’s very sad
He’s writing these authentic love letters, you know, he’s you know, it’s like a service
So he’s like basically writing fake love letters. They’re not really authentic. He’s one writing them for people and very lonely guy
And then he installs this AI operating system and and he you know at first he thinks
Oh, it’s just like that’s kind of cool toy or something and then next thing you know, he’s fallen in love with it
Right and and at some point, you know not to give spoilers, but you know at some point
It kind of outgrows him. Yeah, basically and then if you know at some point it’s like, okay
It’s obviously you need relationships with actual humans
So as a supplement though, I could see this being great for therapy
Like maybe you have a real therapist and then you have like the AI is like, okay
When I can’t talk to the therapist can’t talk to him all the time or what if the therapist’s advice to everybody is just buy more
GPUs Jensen Jensen take it over. You’re wearing his shirt the more you buy the cheaper it gets or whatever his slogan is
The more you buy the more you save money
Yeah, but but you know, also I think about like translation and stuff
So, you know, I recently tweeted about like, you know, I got engaged in Japan and I was using AI to kind of
facilitate the conversation because I can speak a little bit of Japanese but not enough to have like a deep conversation and then
Yeah, without this technology, I wouldn’t have got engaged in like my life’s a lot better now
And I’m sitting there thinking about like, God, what does that got?
You know, how’s that gonna be when you have like AI voice with this too? It’s kind of exciting
It’s kind of weird too. Like I don’t really want the AI voice to be like the main voice she hears from me
Yeah, I’m actually curious, right? How does that interaction look?
I did when you guys have conversations with each other you and your fiance
Yeah, do you like have a phone between you you speak in English and then it says it out
In Japanese and then she says it in Japanese and it speaks it back in like, what does that look like?
I’m just curious. Well, yeah at first it was almost entirely using the phone and occasionally would use Google translate
But like Google Translate’s results are really bad. It almost always has a mistake
But the reason you use Google translate is simplicity, right? It’s faster. You’re not waiting a result
It’s like just it’s right there, but it makes so many mistakes and a few times we had like
Complete misunderstandings because we were using Google translate. I’m like, let me put this in the chat GBT. Oh, okay
Like you said something totally different
You know, I’m not gonna go into details, but there’s once or twice or it was like, oh jeez like
We’re really misunderstanding on something like you know something something big
But you know, but now how it’s kind of evolved
So at first it was like entirely using mostly chat to be especially when we weren’t in person
It was chat to be tea a lot and then in person it was a lot of Google translate because it’s faster
Right and then now how it’s kind of evolved now is you know, I speak a little bit Japanese my son from a previous marriage
He’s half Japanese and so I’ve got a little bit exposure to Japanese and and then she speaks a little bit of
English she loves American movies and things like that, right? So she knows she knows some words, right?
But she just can’t like you know say a whole paragraph or something, right?
So now it’s kind of evolved where in person we mainly use chat GPT for like something really detailed like a long conversation
Like a really deep conversation
We’ll be using AI but for other little things
We’ve like got little words and things where we know, you know, we can communicate basic thing
I use custom instructions with chat GPT to tell it how to teach me as well
Oh, nice like don’t just translate it but actually break down the key words underneath the translation
And so when it would translate for me it’d be like, oh, here’s this word and then here’s a hiragana for it
Which is Japan has like three writing systems and hiragana is the simplest one
They teach kids first and so I’m like show me hiragana cuz I know hiragana
And so it’s not only was it translating but it was it’s been teaching me at the same time
I think I didn’t properly explain that on Twitter cuz people were like, dude, you’re gonna have to learn Japanese
I’m like, but yeah, but it’s been an amazing tool like this relationship would not have happened without that and you know
Not even just in a if you think about like a personal context with relationships
But like business relationships like what is this gonna do for people where now you can go meet a business person in Japan or
China or Saudi Arabia or whatever and and you can use their local language
You probably like have a little device you put down on the table or something. Yeah, you talk
You know, probably like I’m actually spit out something whether it’s in their headphone or whatever
Yeah, it already this is like probably the next six months yet exists and the quality is okay
But like, you know, probably like in the next year will be like really really really good
I was I was at CES back in January and there was a company there called Time Kettle and
Time Kettle has these little earpieces. They just look like like air pods that you’d get from like Apple, right?
That’s kind of what they look like and there’s two of them
I put one in my ear you put one in your ear
I just speak naturally in my language
But what you hear in your earpiece is translated automatically for you and vice versa. So battlefish
So we can just sit there with those those little earpieces in and have a conversation right with in two different languages
Yeah, so I mean that that already exists
You know, you see it at like UN right like when you look at like UN meetings where somebody’s up in front of the whole UN
Speaking everybody in the UN has like one of those little headphones in but I think there it might even actually be a human
Translating for every single person there. I don’t know for sure. That’s my understanding because that is yeah
So I mean eventually I think all of that’s just gonna be like an AI just automatically translating it for whatever language
He turned the dial to you know, yeah, it’s in the book what Hitchhiker’s guide to the galaxy right where they have the
Year
Yeah, it automatically translates it for you like yeah, so we’re heading there like in the next year
Which is gonna be amazing. So yeah, there are products are like niche products
It’s some people know about but there’s not like a mainstream right like amazing automatic translation
You know device or product and I would imagine most people are just gonna want to use the thing
They’ve already got in their pocket, right? Like yeah
Most people probably are gonna go and invest in something like that when they can pull out their phone and just sort of hand it back and forth
I bet there’s like five to ten well-funded startups right now
Working on this. I’d be shocked if there’s not like like it’ll be it’ll be a next wave that you’ll see in the next like six months
To a year like what you did there with the next wave. Yeah, yeah
They’ll be like five of those like all of a sudden, you know, and it’s because yeah
It’s a great idea. Of course somebody’s gonna, you know build that in and win in that market
Well, I think the smartest move that a company can do is for Apple to just build it into the air pod somehow
Right like yeah, everybody already has these little air pods sitting around like just build it into there
Where somebody could speak to me and like it’ll use the processing on my phone, right?
The phone will be in my pocket
Yeah, but these can already hear and they can already produce sound back into my ear, right?
So why not listen to what they say send the information to my phone translate it send it back to my air pods
Like I there’s almost no doubt in my mind that it will eventually just be built into like our earbuds that we use now
That’ll help business so much like you’ll be able to go do business in so many more countries with less
Understandings like right now if you travel around the world like when you go to other countries
They don’t speak your language like yeah sure some places might speak English
But still it’s kind of daunting when you go to a country where they don’t speak English
And now that I’ll be gone like you just put in your ear and you off you go
Yeah, and so that’ll really start to connect the world more
I believe make people understand each other better than people currently do yeah, absolutely the other topic
I want to talk about real quick before we wrap up here is is AI music has had some
Really big advancements within the last few months
I don’t know if you had a chance to play around with like sooner version three yet, but it can make like I haven’t no
It can make up to two minute songs
It actually writes the lyrics creates all the background music and sings it and like the songs are actually good
Like I’ve played some on some of my YouTube videos and people are like I’m actually digging this song
I can listen to that like they’re actually good songs, and it’s just so so so impressive
What they’re doing with with the music now, but also on the flip side of that that coin
200 artists just this week that we’re recording this 200 artists all side of position to try to stop the advancement of AI in the music industry
because it it
It creates an existential threat to their their income their their business model, right?
So like on one hand like the AI music tech is getting so
Unblowing good that we can create whatever song we feel like listening to right now
And it’ll be a unique good interesting song that we like but on the other end of the spectrum
all of the traditional
Musicians and all of the bands that we grew up liking and the current pop artists of today are all fighting hand-in-tooth against it
Yeah, I want to try out the new Suno, so I haven’t tried it yet. I heard about it sounds amazing
I’ll probably try it right when we got here
But yeah, I created something like the top Twitter threads back, you know many many months back now like on a AI Grimes where she had like a
She’s allowing people to use her her voice to make songs
Which I you know that’s kind of going on the one side of the spectrum
Which seems to be pretty rare because everyone else most of the top musicians seem to be like yeah
Don’t use my voice especially without my permission and so like that the other big one that came out was the AI Drake song
And I that that song, you know, I’m a kind of Drake fan like I’m not like like five of his songs
I’m not like a hardcore fan, but I heard that like oh, yeah, this is now one of my top five
Yeah, yeah, yeah, I put it on my my phone. I was listening to it when I went to the gym and everything
I was like this is crazy that this is an AI Drake song
That’s not from him and then all of a sudden, you know
People start having their their Twitter threads taken down the YouTube videos taken down anything that had AI Drake in it was like gone
Yeah, and so it’s like oh wow
Yeah, he like realized that like that’s a big like if there’s a song that’s almost as good as his songs coming out
If there was a smart way for musicians to monetize their voice being trained to the AI systems
All the musicians would be on board but right now
Yeah
There’s no real smart way for them to make money if their voice is being used right but like yeah
It got to a point where I go and create my own variation of a Drake song with lyrics that I created it sounds like Drake
It’s good people like it and
Whenever that music gets played or when this song is generated
I don’t know how the monetization would work
But if there’s a way that yeah Drake made some money every single time that song got played
He would be all for it because now yeah
You can make a million Drake songs can be made and he’s making money off of all of them
But the problem is right these artists have no way of actually making money off of their voice being trained in there
And I think a lot of these companies want to figure that out because if they can crack that code
Yeah, how do we actually incentivize musicians to be a part of this?
Musicians will probably be a lot more likely to be involved in it
Yeah, I kind of feel like we’re gonna probably need like some future AI to help us figure out how to do that
Yeah, you know you probably you probably don’t even know this but like my last startup binded
That’s like kind of what we were going after so we were doing
Attribution on the blockchain and trying to automate royalties and things like that so we experimented with music
We end up doing images because like it was the easiest way to get started
I went out to Washington DC and met with people in the copyright office
I spoke on a panel in Washington DC with the copyright office and
And man, it is like hard. It’s very hard to like track these things and make sure it’s actually authentic and then handle the payments
You’ve got Spotify right and Spotify
They’ve got some sort of model where every time a song gets played the musician gets it like a fraction of a penny or something
But if you’re a popular musician and getting millions and millions and millions of downloads a month
It adds up they can make a living off of it
And I feel like a lot of people are sort of looking at Spotify is like we got to do something like that
But I feel like it’s so much more complicated than what Spotify is trying to do because if you train all these voices
Into an AI and then somebody goes and creates a song, you know, how do you know exactly which voices?
It’s pulling from to create this song
Maybe the voice is a blend between Drake’s voice and Eminem’s voice and it’s like a hybrid now
Do they get paid for that? Like there’s so much more like intricacies involved
And do they get paid and does the publisher and all these other people get paid like who you know who all gets paid?
And then like existing contracts they already have that may like prohibit those kind of things. Honestly, I think it’s just gonna require new contracts
Like a rethinking of that industry essentially, right? Like the music industry already had to reinvent itself when streaming came along, right?
Everybody bought CDs everybody bought albums now do bands even make albums anymore or do they just drop songs, right?
Because like the music industry’s train changed so much
I just feel like they’ve got to figure out what what is this next evolution because the change is going to happen
It’s not like them putting out like 200 of them signing a letter is gonna stop anything
It’s just sort of making their feelings known, but it’s not gonna stop anything. Yeah, man
It’s kind of like with like Napster like yeah, you know, they tried to stop Napster
But that kind of technology and Torrance everything else kept evolving and like, you know, it’s hard to start technology, right?
And so yeah, when there when there’s a 10,000 Drake songs out there AI Drake songs like what are they gonna do like, you know
A million AI Drake songs. Yeah, are they really gonna be able to stop all of that?
Am I gonna go pay to listen to a Drake song if I can use a tool and just generate a new Drake song like that?
I’ve never heard before like right and I do wonder if it’s gonna lead to a world, too
We’re like, you know, you you almost like freeze culture in place, which I hope does not happen
I hope you know culture keeps advancing and there’s new creative works
I hope we don’t like freeze culture in place where it’s like, okay, you’ve got the Beatles you got Michael Jackson
You got whoever and like you’re like replicating their voice to make new songs and like that’s like the famous songs for all of time
No, I think it’ll keep evolving
But I do think AI is just gonna be another tool in the mix, right?
I think I think people are gonna figure out creative ways to use AI
I think AI can be a great tool for like
Musicians to collaborate with other musicians without the other musician need to be involved like maybe
Eminem goes in like produces a song and he wants, you know Drake to cameo on the song and
Drake just doesn’t have the time well Drake could license his weights to Eminem
Eminem can generate the clips that he needs from Drake work them into his song
Yeah, Drake gets paid Eminem just Drake on his song and they never had to meet up in person, right?
I think there’s something there if they could just figure out all the logistics of it
Yeah, and like bringing it back to AI voice and you could be like mixing all of this with your voice, too
Right, you’d be like, yeah changes part of the song or add something here add Drake in this part and like doing a lot of that
With voice that’s gonna be you know, that’s gonna be a new creative experience
Which which probably would be great because like people can say more in the flow, right?
It would be more creative like without having to like go manually touch all these tools
Like you just kind of making the music and like, you know going with the flow
So now that we’ve talked about AI voice and AI music and where all of this is headed you probably want to know
All right, what are what are the takeaways?
What what can I do with this information and there’s a few things when it when it comes to like the risks and dangers of AI
One of the things I actually told my parents is that if they ever get a call from me or my wife or one of my kids
We have a code word ask for that code word to make sure it’s really us
I mean if I’m just called to say happy birthday
You don’t ask to ask ask for the code word
But if I’m asking for money if I’m you know if if I’m saying there’s a problem where I was in a car accident and
I need money or somebody was kidnapped and I need money or something that sounds really really out of the ordinary
Ask for me ask me for the code word to verify that it’s really me because that is
Something that I think people should start doing because AI voice is only gonna get better and better
So I think that’s like one of the things that you should really
Really take away and another thing is I think you should go and use a lot of these tools
I think you should try 11 labs. You should listen to the open AI voice stuff
You should listen to the sooner music. I think the more you get immersed like we’ve talked about
The deeper we go down these AI rabbit holes the better we get at detecting whether this is AI or real
It’s almost like you know back when Photoshop came out people had a hard time telling whether something was photoshopped or not
But over time you see enough of it and now people can go okay. That looks like it was probably photoshopped
I feel like the same kind of thing can happen with AI audio over time. You’ll probably get better and better at
Detecting AI. That’s not to say AI is not gonna get better and at some point it will be undetectable
But short term you should probably be using these tools hearing them
Understanding how they sound and you will probably get better and better at seeing these little nuances or hearing these little nuances in the audio
They give away that it’s that it’s made with AI
Well, and I think the other key takeaway too though is that like a lot of the like
Circle all the way back around bookend it to how it started of like the other key takeaway is that
All of this is going to turn to voice as opposed to typing one thing people need to realize is you know
This technology is not just like a sci-fi movie like her now
It’s you know
It’s here now and you’ll probably see in the next six months to 12 months that the main way people are using AI is
Voice and so you know as a as a business leader executive
Employee you should be thinking about how are you gonna be able to use these tools in a year from now with voice that you can’t currently with
Text and and how is your how is your life gonna look differently when you can just talk to the AI and have it help do work for you?
You know, you’ll even be able to do things like you know
Create AI agents where you’ll be able to send off the agents to do a little task for you and command them by voice
Right, so like that’s coming very very soon
So like imagine, you know be planning for that and you’ll be in a way better position
Than people who have no idea this technology exists and if you have a business that has content online
It’s really easy these days to make an audio version of that written content as well
So I think a lot more people are going to also consume content via audio, right?
I think the prompting is gonna be more audio-based where we talked to Siri
We talked to Alexa we talked to these tools and it sort of goes and does the prompt based on what we say to it
But I also think the reverse is true where over time more and more people might consume their content that way as well
So if you have a blog with written content
Throw that content into 11 labs and to have it a podcast audio version that people can listen to as well
Because now you just have another format that makes it more likely that somebody’s gonna consume the content that you just created
So I think that’s another like take away for businesses listening to this is lean into this use this technology
It’s actually a really cool way to make audio versions of your content
Yeah, and I think you know in terms of you know people should be playing too for like how they can use this internationally, right?
Like this is gonna open up so many opportunities that your business can’t currently take advantage of
You know think about like okay people who speak different languages
I’ll know be able to have business meetings with them or think about if you’re making videos or written content that you didn’t turn into audio
You’ll be able to turn that into like a hundred different languages
Right like what does that mean for your business?
So I got so everyone should be thinking about that right now and hopefully we’ll do the same with the podcast
Hopefully we’ll have this in Japanese the next six months or something like that brings up another question when you proposed
Did you actually propose with the cell phone like a translator?
I tried to not use it and then it was it was it was necessary within 30 seconds
But I tried my best and another thing is there the other thing I didn’t mention is you know my son
He’s bilingual. He speaks English in Japanese both perfectly early or at least for a 10 year
You know a 10 year old and and so occasionally, you know
He helps translate which I try to make sure he doesn’t do that too often. It’s like kind of you know
Annoying for him, but but but he but he was there as well like in another room. So I was like, well worst-case scenario
I’ll be like, you know, no
Very cool. Well AI is is changing lives and building relationships. Yeah
Exciting times we’re in bringing loved everyone
And yeah, okay go watch the movie her I’ll go watch the movie her and on that note
I think we can wrap this one up. Awesome. Well, thank you so much to everybody for tuning in
Please like this video and subscribe to our channel if you haven’t already it really really helps get our podcast in front
Of more people if there’s somebody that you know that this episode can be helpful for send them the link
Let them let them tune into this episode if you’re on a podcast player like Spotify or Apple
Give us a subscribe and maybe even leave us a review. We really really appreciate it
It helps spread the word of this podcast and shares this information with more people. So thanks again for tuning in
You
You
You
You
(chiming music)
[BLANK_AUDIO]
People are using AI is voice
Hey, welcome to the next wave podcast my name is Matt Wolf
I’m here with my co-host Nathan Land and we are your chief AI officer
It is our goal with this podcast to keep you looped in on all the latest AI news the coolest AI tools and
Set you and your business up to be ready for this next wave of AI that’s coming and today
We’re going to talk about what we think is actually the next wave of AI your voice in this episode
We’re gonna talk about why we think voice is the next big thing
We’re gonna talk about the tools that are available so that you can use voice
We’re gonna talk about the scary risks and why this could be fairly damaging to the world if things go in the wrong direction
we’re also gonna talk about how AI is taking over the music world and
What potential options are out there to solve some of these problems?
so a really really fascinating episode lots of cool stuff we’re gonna talk about and
Excited to share with you. So let’s go ahead and just dig in
Yeah, so today when everyone thinks about AI most people are thinking about text using text to talk to AI
You know like chat GBT Claude
but I think the big a-ha moment for a lot of people is gonna be that probably in the next six months to a year the
Main way people are gonna interact. I believe is gonna be with voice, you know
And this company Hume AI just raised 50 million dollars in their series B
And they look they just launched a product called Evie, which is a empathetic voice interface and you know
I tried it yesterday with my son and
And he was blown away, you know, he’s ten years old
He tried it out and he was like what is that and he saw like the little line on it where it’s showing the emotions
You know and he started and he started talking to it and he was just like freaking out because it was
Detecting the emotions in his voice and then responding accordingly
When all your marketing team does is put out fires they burn out fast
Sifting through leads creating content for infinite channels endlessly searching for disparate performance KPIs
It all takes a toll but with HubSpot you can stop team burnout in its tracks
Plus your team can achieve their best results without breaking a sweat with HubSpot’s collection of AI tools
Breeze you can pinpoint the best leads possible capture prospects attention with click-worthy content and
Access all your company’s data in one place no sifting through tabs necessary
It’s all waiting for your team in HubSpot
Keep your marketers cool and make your campaign results hotter than ever visit hubspot.com slash marketers to learn more
No, it’s really interesting. I played around with you
I actually made a YouTube video where it was like the AI news recap
But in one piece of it, I played around with Hume for a minute
And it was kind of funny because I was like hey Hume, how you doing today or something like that?
And it was like I I detect that you’re angry. Why are you angry and I’m like I’m not angry
I think I was probably like when I make videos and when we record podcasts like this
I tend to ramp up my energy and you know
I’m probably a little bit more excited than if I was you know
Just talking to you in real life face-to-face and I have a feeling Hume sort of picked up on that sort of extra
Excited energy news like what is wrong with you? Yeah. Yeah. Yeah. Yeah. Yeah, but anyway, you know
Back to the bigger point here is that I do feel like we’re entering this world where right now
Most of the prompting we do is with text prompting you go to chat gpt
You go to Claude you go to one of these platforms right and you type in what you want it to respond to
But I feel like over the last decade or so the tech world has kind of been training us that know your voice should be
Mechanism for interacting with the computer right with things like Siri and what’s Google’s version Google’s assistant version
Is it not Google home? Yeah, but anyway the point being like yeah, we have all of these smart devices
We’ve got our phones. We’ve got our Alexa’s. We’ve got stuff like that
That is I think where things are going right you’re going to be actually speaking to these things instead of typing prompts
and with things like Hume, we’re gonna actually get even more context for it to analyze when it gives us feedback, right?
The rumor has it that at WWDC this year Apple’s big developer conference that they do every year
That’s kind of where they make the big announcements
It was WWDC last year where they announced Apple Vision Pro this year
The big expectation is going to be a lot of AI right and I think a lot of people are speculating that Siri is going to get AI
Into it what they’re gonna use for the AI. We don’t really know yet
You know, we we do know that there’s been rumors going around that Apple and Google might partner and that Apple might use Gemini
1.5 for Siri behind the scenes, but then also
Apple just released a research paper the week that we’re recording this apple just released a research paper called realm
And it’s basically their large language model that they designed for a mobile phone. It is actually very good at reading the context of what’s on the screen
it can actually view a mobile phone screen and
Use that context to sort of inform the large language model before it responds to you
So a lot of people are thinking well, maybe that’s what’s gonna be in Siri, right?
So maybe you’re on an app. You’re using the app. You might be able to go Siri. Hey, this app isn’t working
Can you help troubleshoot it for me? Siri will actually be able to see your iPhone screen see what’s going on and
Possibly help you through whatever issue you’re running into if they use this realm language model that Apple just put out the research for
But where they actually go with it. I don’t think anybody’s gonna actually know until WWDC, which is I believe in June of this year
I think it’s crazy that like, you know, Alexa and Siri both came out like 10 or 11 years ago right around the time when the movie
Her came out, you know, everyone thought this technology was gonna, you know, improve every year. It was gonna be amazing
And so a lot of people I think they think oh this stuff is not gonna get better because look look it took 10 years
It’s still almost the same product
But they don’t realize like when like GPT-5 comes out if you built something like Alexa now with like GPT-5 under the hood
How dramatically different of an experience that’s that’s gonna be and and I’m sure Alexa will be upgrading soon too because you know
Amazon just invested. What was it like four billion dollars and then something like that. Yeah
Well, they initially invested like 1.75 billion or something and then when yeah all this data came out that
Opus is actually beating the latest version of GPT-4 in all of the benchmarks
Yeah, Amazon came back to Anthropic and said hey, maybe we’ll put even more money into you guys
And then I think they invested like another two and a half billion or something like that, right?
So we’re gonna have Jarvis in our house pretty soon if you want to yeah
And if you think of that from like a business context, so you know as a CEO or a manager like how amazing is that gonna be that instead of
Having to type everything to your AI you can be like really quickly like hey
You know this is going on or hey, can you check this report for me or can you check?
You know what do I have an email from so-and-so and can you respond to that doing all of that by voice like how much more
Efficient you’ll be as an executive using these kind of tools. Yeah
While we’re on the topic of voice to open AI just put out some new research as well, right?
Open AI just put out an article on their blog saying that we’ve got this really really good voice model where with I don’t remember what it
Was exactly but like 15 seconds of training data. We can replicate your voice and then you can type in whatever you want
It’ll say that thing. Yeah, but then open AI in that same article wet, but it’s too powerful guys. It’s too strong
It’s too scary. We can’t let you use that yet, which to me is just wild because we’ve already got it, right?
We’ve got 11 labs 11 labs already do exactly that right we’ve had like uber doc
There’s been there’s been a whole bunch of models out there already that do exactly that an open AI kind of coming out and saying like
We do it too, but we’re not giving you our version because it’s too good kind of weird
So when I saw Evie from Hume, I was like, okay, that’s probably like the best I had seen so far like on the emotional side
That’s it’s amazing to texture emotions, but you could still kind of tell it’s not a human and with 11 labs pretty good
But you can still kind of tell it’s not a human. I know that the demo I saw I believe it’s from open AI
That was the most realistic one I had heard. I was like, okay, that’s like 99.5 percent there
Yeah, like if I didn’t know this technology would existed. I would not know that was AI
I would think that was a real person talking. Yeah. Yeah, the other ones are like 95 percent there
But when you get to like 99 like people just they don’t know that it’s an AI and that’s where it’s you know
I can get that. Yeah, that could be kind of scary that you could just be pretending to be anyone
Yeah, no one would know well totally
But I also think that you know people like you and I who are immersed in it on a daily basis are probably a little more
Adepted actually spotting AI right like I feel like I’m really good at when it comes to AI images and going I can tell like
Instantly that it’s AI now because I’ve seen so many AI images
I feel like I’m getting that way with yeah voice now to where I can pretty quickly grasp
Okay, that was done with AI
But I mean how many people have actually been fooled already by AI voice using things like 11 labs
I mean that that that kind of sort of brings us into the I guess kind of the next thing that I want to talk about is like a
Scary part about all of this AI voice technology that’s that’s coming out right now is you know
there’s there’s been stories of a
Scammer calling up somebody’s parents and saying hey
We kidnapped your daughter and then they would use a voice sample of that person
to convince the person on the other side of the phone like oh they really have my daughter and
Collect ransom money right. There’s also that what was it the 25 million dollar scam that happened over zoom
Like remember when we talked about that for a little bit
Yeah, yeah, like so I mean that yeah stuff’s happening right now
Yeah, and I think that was here in Japan. I believe and so yeah somebody
impersonated an executive or I think they print impersonated with the CFO or was like a top executive at a big company and
Called people up and people didn’t know that it wasn’t a real person
And they end up wiring. What was it 25 million or something or something like that?
It was a lot of money that they just wired off to some scammer. Yeah
Yeah, the tech to do the trickery is already out there
Yeah, but like you said like yeah open AI is the best one to come along so far
Yeah, but it’s already out there like people can already do this right now with it with this kind of technology
So there’s already laws that this this kind of stuff is illegal like so I guess the challenge is the scale will probably greater
Because like it’ll be easier to scam people now. Yeah, right. So and that is probably an area where you’ll need AI to help like
You know find scammers and things like that, right?
Which you know hopefully doesn’t go into like big brother territory where we’re using AI like monitor everyone
But you know possibly that’s it is slightly going towards that direction like yeah
You need AI to like figure out like so-and-so scamming people and then yeah look at the data and then go get the person
Yeah, I’ve got somebody that like is anti regulation. I actually think regulation to some degree is of a good thing
In the sense that I feel like regulation gives companies bumpers to stay within I think that could be a good thing
Right, I think right now. There’s a lot of AI companies out there. They’re developing and they’re going well
We don’t know what’s gonna happen with regulation
So let’s just keep on pushing the limits
But I do feel like regulation to some degree gives them some bumpers to stay within so they know they’re not overstepping their bounds
However, when it comes to AI, I don’t feel like regulation works at all
Because bad actors are gonna be bad actors, right? So we can go out there and say like hey
We’re gonna regulate you’re not allowed to clone people’s voices with AI and do people
Well, you already can’t do that, right? Whether it’s law or not, you know, it’s already highly unethical people already see that as a negative thing
I’m pretty sure it’s already a law that you can’t do that anyway
But yeah, yeah, the technology’s out there. It’s out there in open source form. It’s out there in closed source form
It’s out there people are gonna be able to do it. So how do you like what is regulation going to accomplish in that sense?
I don’t really understand the regulation argument here
You know, most of the politicians have very little like real-life business experience
And so like when they think about regulation a lot of times it’s just like, okay, is the public scared of this technology?
Well, okay, then I’m gonna regulate it. It’s like it’s actually good for your country. Is it good for like business?
Is it good, you know, are you just literally responding to some polls? Is that all that is and you know without going too deep down like the sort of
Political rabbit hole there’s lobbyists, right? And a lot of the government is run by the companies who are paying
For those people to be where they are in the government, right?
So what I think what the big fear is and I know this is something whether you love them or hate them that Gary Marcus has talked about a
Little bit in the past. Yeah, is that what happens when we get too much regulation with this is what it’s gonna end up
Doing is concentrating the power into a few small companies, right?
What’s gonna end up happening is the the companies like open AI and Microsoft and maybe Google and some of these big corporations that are
Pushing for the regulations may end up driving the regulations in ways that really favor their company
But really do not favor the little guys, you know, that’s kind of like the biggest divide that’s happened in Silicon Valley recently
Is like you got the divide between like the Sam Altman, you know side
Which is also kind of aligned with the YC and then the Mark Andreessen VC side where they’re like, yeah
This is basically regulatory capture here. Like they’re like trying to go out there and like say yeah
What we’re building is very dangerous. So you should please regulate. Yeah, that’s an odd thing to ask for like why are they doing that?
Please regulate us and we’ll be on the board to decide what those regulations are
Yeah, yeah, yeah regulate us in this exact way that we want that no startup can afford
Yeah, and oh by the way, open source is very dangerous and so that’s kind of been like the undertones of it all too
It’s like when they’re talking about regulation. It’s like open sourcing of AI is very dangerous
That’s almost always the undertone. It’s like well, if you don’t have open source AI
Then yeah, you will end up with like one or two companies that controls all the technology
So so that’s why I’m like very hesitant of like like I’m not gonna say there’s no regulation
They did like maybe there are there’s regulation needed for like, you know deep fakes or you know AI voice at some point
But like just speeding into it and trying to regulate everything kind of like what’s going on and you’re a little bit right now
I’m like, that’s not the right approach
Hey, you know a lot of a lot of where my heads at to kind of comes from like the crypto space, right?
Where there’s been looming regulations forever, but the a lot of the regulations never end up happening
So a lot of the companies that are trying to build or like am I gonna build something that’s gonna end up getting regulated out of existence?
So like a lot of people and companies in that crypto space are like just tell us the damn regulations
So we know what bumpers to stick with it and we’re not in this limbo
But when we’re like talking about AI, I feel like it’s a different story because it’s like you do have the open source
You do have the closed source
You do have the companies that you know have a better foothold within
Governments than the little guys you do have all these other nuances
I think really muddy the waters and also I feel like regulation is just gonna make it easier for the bad actors to be the bad actors
While the the people they’re trying to do right lose abilities essentially right along with all the negatives that we just talked about
There are still some positives of this this voice AI that’s coming out
I think there are some use cases that I think could be very valuable
I know for me as a content creator using it for like dubbing is really really helpful like if I misspeak in one of my videos
I don’t have to go back in record and like overdub something
I can open up a tool like descriptor 11 labs type in the words that I meant to say and then just use that and dub it into my
Video so I’ve actually used some of these tools to sort of fix a misspeak in some of my videos
So that’s like that’s one really good use case for some of this AI voice
We’ll be right back but first I want to tell you about another great podcast you’re gonna want to listen to it’s called science of scaling
Hosted by mark robert’s and it’s brought to you by the hub spot podcast network the audio
Destination for business professionals each week host mark robert’s
Founding chief revenue officer at hub spot senior lecturer at Harvard Business School and co-founder of stage two capital
Sits down with the most successful sales leaders in tech to learn the secrets strategies and tactics to scaling your company’s growth
He recently did a great episode called how do you solve for a siloed marketing and sales
And I personally learned a lot from it. You’re gonna want to check out the podcast
Listen to science of scaling wherever you get your podcasts
Yeah, and human they’re kind of pitching that this is gonna be used for therapy and things like that right which is
Exciting you know, it’s kind of you know gonna be weird to talk to a robot
And that’s how you’re getting your therapy, but I mean who knows like when it gets good enough
You know, I guess that’ll be a thing it does kind of make me think of the movie her
Yeah, right, you know, we’re you’ve got this guy who I think you haven’t seen it
And I feel like you should you should take my AI fan card from me. Yeah, you seem ex Machina. I have seen ex Machina. Yes
Okay, okay. Okay. You’re okay. You’re like you’re halfway there, you know
So with her you’ve got you know, Joaquin Phoenix playing this guy who’s like got divorced. He’s very sad
He’s writing these authentic love letters, you know, he’s you know, it’s like a service
So he’s like basically writing fake love letters. They’re not really authentic. He’s one writing them for people and very lonely guy
And then he installs this AI operating system and and he you know at first he thinks
Oh, it’s just like that’s kind of cool toy or something and then next thing you know, he’s fallen in love with it
Right and and at some point, you know not to give spoilers, but you know at some point
It kind of outgrows him. Yeah, basically and then if you know at some point it’s like, okay
It’s obviously you need relationships with actual humans
So as a supplement though, I could see this being great for therapy
Like maybe you have a real therapist and then you have like the AI is like, okay
When I can’t talk to the therapist can’t talk to him all the time or what if the therapist’s advice to everybody is just buy more
GPUs Jensen Jensen take it over. You’re wearing his shirt the more you buy the cheaper it gets or whatever his slogan is
The more you buy the more you save money
Yeah, but but you know, also I think about like translation and stuff
So, you know, I recently tweeted about like, you know, I got engaged in Japan and I was using AI to kind of
facilitate the conversation because I can speak a little bit of Japanese but not enough to have like a deep conversation and then
Yeah, without this technology, I wouldn’t have got engaged in like my life’s a lot better now
And I’m sitting there thinking about like, God, what does that got?
You know, how’s that gonna be when you have like AI voice with this too? It’s kind of exciting
It’s kind of weird too. Like I don’t really want the AI voice to be like the main voice she hears from me
Yeah, I’m actually curious, right? How does that interaction look?
I did when you guys have conversations with each other you and your fiance
Yeah, do you like have a phone between you you speak in English and then it says it out
In Japanese and then she says it in Japanese and it speaks it back in like, what does that look like?
I’m just curious. Well, yeah at first it was almost entirely using the phone and occasionally would use Google translate
But like Google Translate’s results are really bad. It almost always has a mistake
But the reason you use Google translate is simplicity, right? It’s faster. You’re not waiting a result
It’s like just it’s right there, but it makes so many mistakes and a few times we had like
Complete misunderstandings because we were using Google translate. I’m like, let me put this in the chat GBT. Oh, okay
Like you said something totally different
You know, I’m not gonna go into details, but there’s once or twice or it was like, oh jeez like
We’re really misunderstanding on something like you know something something big
But you know, but now how it’s kind of evolved
So at first it was like entirely using mostly chat to be especially when we weren’t in person
It was chat to be tea a lot and then in person it was a lot of Google translate because it’s faster
Right and then now how it’s kind of evolved now is you know, I speak a little bit Japanese my son from a previous marriage
He’s half Japanese and so I’ve got a little bit exposure to Japanese and and then she speaks a little bit of
English she loves American movies and things like that, right? So she knows she knows some words, right?
But she just can’t like you know say a whole paragraph or something, right?
So now it’s kind of evolved where in person we mainly use chat GPT for like something really detailed like a long conversation
Like a really deep conversation
We’ll be using AI but for other little things
We’ve like got little words and things where we know, you know, we can communicate basic thing
I use custom instructions with chat GPT to tell it how to teach me as well
Oh, nice like don’t just translate it but actually break down the key words underneath the translation
And so when it would translate for me it’d be like, oh, here’s this word and then here’s a hiragana for it
Which is Japan has like three writing systems and hiragana is the simplest one
They teach kids first and so I’m like show me hiragana cuz I know hiragana
And so it’s not only was it translating but it was it’s been teaching me at the same time
I think I didn’t properly explain that on Twitter cuz people were like, dude, you’re gonna have to learn Japanese
I’m like, but yeah, but it’s been an amazing tool like this relationship would not have happened without that and you know
Not even just in a if you think about like a personal context with relationships
But like business relationships like what is this gonna do for people where now you can go meet a business person in Japan or
China or Saudi Arabia or whatever and and you can use their local language
You probably like have a little device you put down on the table or something. Yeah, you talk
You know, probably like I’m actually spit out something whether it’s in their headphone or whatever
Yeah, it already this is like probably the next six months yet exists and the quality is okay
But like, you know, probably like in the next year will be like really really really good
I was I was at CES back in January and there was a company there called Time Kettle and
Time Kettle has these little earpieces. They just look like like air pods that you’d get from like Apple, right?
That’s kind of what they look like and there’s two of them
I put one in my ear you put one in your ear
I just speak naturally in my language
But what you hear in your earpiece is translated automatically for you and vice versa. So battlefish
So we can just sit there with those those little earpieces in and have a conversation right with in two different languages
Yeah, so I mean that that already exists
You know, you see it at like UN right like when you look at like UN meetings where somebody’s up in front of the whole UN
Speaking everybody in the UN has like one of those little headphones in but I think there it might even actually be a human
Translating for every single person there. I don’t know for sure. That’s my understanding because that is yeah
So I mean eventually I think all of that’s just gonna be like an AI just automatically translating it for whatever language
He turned the dial to you know, yeah, it’s in the book what Hitchhiker’s guide to the galaxy right where they have the
Year
Yeah, it automatically translates it for you like yeah, so we’re heading there like in the next year
Which is gonna be amazing. So yeah, there are products are like niche products
It’s some people know about but there’s not like a mainstream right like amazing automatic translation
You know device or product and I would imagine most people are just gonna want to use the thing
They’ve already got in their pocket, right? Like yeah
Most people probably are gonna go and invest in something like that when they can pull out their phone and just sort of hand it back and forth
I bet there’s like five to ten well-funded startups right now
Working on this. I’d be shocked if there’s not like like it’ll be it’ll be a next wave that you’ll see in the next like six months
To a year like what you did there with the next wave. Yeah, yeah
They’ll be like five of those like all of a sudden, you know, and it’s because yeah
It’s a great idea. Of course somebody’s gonna, you know build that in and win in that market
Well, I think the smartest move that a company can do is for Apple to just build it into the air pod somehow
Right like yeah, everybody already has these little air pods sitting around like just build it into there
Where somebody could speak to me and like it’ll use the processing on my phone, right?
The phone will be in my pocket
Yeah, but these can already hear and they can already produce sound back into my ear, right?
So why not listen to what they say send the information to my phone translate it send it back to my air pods
Like I there’s almost no doubt in my mind that it will eventually just be built into like our earbuds that we use now
That’ll help business so much like you’ll be able to go do business in so many more countries with less
Understandings like right now if you travel around the world like when you go to other countries
They don’t speak your language like yeah sure some places might speak English
But still it’s kind of daunting when you go to a country where they don’t speak English
And now that I’ll be gone like you just put in your ear and you off you go
Yeah, and so that’ll really start to connect the world more
I believe make people understand each other better than people currently do yeah, absolutely the other topic
I want to talk about real quick before we wrap up here is is AI music has had some
Really big advancements within the last few months
I don’t know if you had a chance to play around with like sooner version three yet, but it can make like I haven’t no
It can make up to two minute songs
It actually writes the lyrics creates all the background music and sings it and like the songs are actually good
Like I’ve played some on some of my YouTube videos and people are like I’m actually digging this song
I can listen to that like they’re actually good songs, and it’s just so so so impressive
What they’re doing with with the music now, but also on the flip side of that that coin
200 artists just this week that we’re recording this 200 artists all side of position to try to stop the advancement of AI in the music industry
because it it
It creates an existential threat to their their income their their business model, right?
So like on one hand like the AI music tech is getting so
Unblowing good that we can create whatever song we feel like listening to right now
And it’ll be a unique good interesting song that we like but on the other end of the spectrum
all of the traditional
Musicians and all of the bands that we grew up liking and the current pop artists of today are all fighting hand-in-tooth against it
Yeah, I want to try out the new Suno, so I haven’t tried it yet. I heard about it sounds amazing
I’ll probably try it right when we got here
But yeah, I created something like the top Twitter threads back, you know many many months back now like on a AI Grimes where she had like a
She’s allowing people to use her her voice to make songs
Which I you know that’s kind of going on the one side of the spectrum
Which seems to be pretty rare because everyone else most of the top musicians seem to be like yeah
Don’t use my voice especially without my permission and so like that the other big one that came out was the AI Drake song
And I that that song, you know, I’m a kind of Drake fan like I’m not like like five of his songs
I’m not like a hardcore fan, but I heard that like oh, yeah, this is now one of my top five
Yeah, yeah, yeah, I put it on my my phone. I was listening to it when I went to the gym and everything
I was like this is crazy that this is an AI Drake song
That’s not from him and then all of a sudden, you know
People start having their their Twitter threads taken down the YouTube videos taken down anything that had AI Drake in it was like gone
Yeah, and so it’s like oh wow
Yeah, he like realized that like that’s a big like if there’s a song that’s almost as good as his songs coming out
If there was a smart way for musicians to monetize their voice being trained to the AI systems
All the musicians would be on board but right now
Yeah
There’s no real smart way for them to make money if their voice is being used right but like yeah
It got to a point where I go and create my own variation of a Drake song with lyrics that I created it sounds like Drake
It’s good people like it and
Whenever that music gets played or when this song is generated
I don’t know how the monetization would work
But if there’s a way that yeah Drake made some money every single time that song got played
He would be all for it because now yeah
You can make a million Drake songs can be made and he’s making money off of all of them
But the problem is right these artists have no way of actually making money off of their voice being trained in there
And I think a lot of these companies want to figure that out because if they can crack that code
Yeah, how do we actually incentivize musicians to be a part of this?
Musicians will probably be a lot more likely to be involved in it
Yeah, I kind of feel like we’re gonna probably need like some future AI to help us figure out how to do that
Yeah, you know you probably you probably don’t even know this but like my last startup binded
That’s like kind of what we were going after so we were doing
Attribution on the blockchain and trying to automate royalties and things like that so we experimented with music
We end up doing images because like it was the easiest way to get started
I went out to Washington DC and met with people in the copyright office
I spoke on a panel in Washington DC with the copyright office and
And man, it is like hard. It’s very hard to like track these things and make sure it’s actually authentic and then handle the payments
You’ve got Spotify right and Spotify
They’ve got some sort of model where every time a song gets played the musician gets it like a fraction of a penny or something
But if you’re a popular musician and getting millions and millions and millions of downloads a month
It adds up they can make a living off of it
And I feel like a lot of people are sort of looking at Spotify is like we got to do something like that
But I feel like it’s so much more complicated than what Spotify is trying to do because if you train all these voices
Into an AI and then somebody goes and creates a song, you know, how do you know exactly which voices?
It’s pulling from to create this song
Maybe the voice is a blend between Drake’s voice and Eminem’s voice and it’s like a hybrid now
Do they get paid for that? Like there’s so much more like intricacies involved
And do they get paid and does the publisher and all these other people get paid like who you know who all gets paid?
And then like existing contracts they already have that may like prohibit those kind of things. Honestly, I think it’s just gonna require new contracts
Like a rethinking of that industry essentially, right? Like the music industry already had to reinvent itself when streaming came along, right?
Everybody bought CDs everybody bought albums now do bands even make albums anymore or do they just drop songs, right?
Because like the music industry’s train changed so much
I just feel like they’ve got to figure out what what is this next evolution because the change is going to happen
It’s not like them putting out like 200 of them signing a letter is gonna stop anything
It’s just sort of making their feelings known, but it’s not gonna stop anything. Yeah, man
It’s kind of like with like Napster like yeah, you know, they tried to stop Napster
But that kind of technology and Torrance everything else kept evolving and like, you know, it’s hard to start technology, right?
And so yeah, when there when there’s a 10,000 Drake songs out there AI Drake songs like what are they gonna do like, you know
A million AI Drake songs. Yeah, are they really gonna be able to stop all of that?
Am I gonna go pay to listen to a Drake song if I can use a tool and just generate a new Drake song like that?
I’ve never heard before like right and I do wonder if it’s gonna lead to a world, too
We’re like, you know, you you almost like freeze culture in place, which I hope does not happen
I hope you know culture keeps advancing and there’s new creative works
I hope we don’t like freeze culture in place where it’s like, okay, you’ve got the Beatles you got Michael Jackson
You got whoever and like you’re like replicating their voice to make new songs and like that’s like the famous songs for all of time
No, I think it’ll keep evolving
But I do think AI is just gonna be another tool in the mix, right?
I think I think people are gonna figure out creative ways to use AI
I think AI can be a great tool for like
Musicians to collaborate with other musicians without the other musician need to be involved like maybe
Eminem goes in like produces a song and he wants, you know Drake to cameo on the song and
Drake just doesn’t have the time well Drake could license his weights to Eminem
Eminem can generate the clips that he needs from Drake work them into his song
Yeah, Drake gets paid Eminem just Drake on his song and they never had to meet up in person, right?
I think there’s something there if they could just figure out all the logistics of it
Yeah, and like bringing it back to AI voice and you could be like mixing all of this with your voice, too
Right, you’d be like, yeah changes part of the song or add something here add Drake in this part and like doing a lot of that
With voice that’s gonna be you know, that’s gonna be a new creative experience
Which which probably would be great because like people can say more in the flow, right?
It would be more creative like without having to like go manually touch all these tools
Like you just kind of making the music and like, you know going with the flow
So now that we’ve talked about AI voice and AI music and where all of this is headed you probably want to know
All right, what are what are the takeaways?
What what can I do with this information and there’s a few things when it when it comes to like the risks and dangers of AI
One of the things I actually told my parents is that if they ever get a call from me or my wife or one of my kids
We have a code word ask for that code word to make sure it’s really us
I mean if I’m just called to say happy birthday
You don’t ask to ask ask for the code word
But if I’m asking for money if I’m you know if if I’m saying there’s a problem where I was in a car accident and
I need money or somebody was kidnapped and I need money or something that sounds really really out of the ordinary
Ask for me ask me for the code word to verify that it’s really me because that is
Something that I think people should start doing because AI voice is only gonna get better and better
So I think that’s like one of the things that you should really
Really take away and another thing is I think you should go and use a lot of these tools
I think you should try 11 labs. You should listen to the open AI voice stuff
You should listen to the sooner music. I think the more you get immersed like we’ve talked about
The deeper we go down these AI rabbit holes the better we get at detecting whether this is AI or real
It’s almost like you know back when Photoshop came out people had a hard time telling whether something was photoshopped or not
But over time you see enough of it and now people can go okay. That looks like it was probably photoshopped
I feel like the same kind of thing can happen with AI audio over time. You’ll probably get better and better at
Detecting AI. That’s not to say AI is not gonna get better and at some point it will be undetectable
But short term you should probably be using these tools hearing them
Understanding how they sound and you will probably get better and better at seeing these little nuances or hearing these little nuances in the audio
They give away that it’s that it’s made with AI
Well, and I think the other key takeaway too though is that like a lot of the like
Circle all the way back around bookend it to how it started of like the other key takeaway is that
All of this is going to turn to voice as opposed to typing one thing people need to realize is you know
This technology is not just like a sci-fi movie like her now
It’s you know
It’s here now and you’ll probably see in the next six months to 12 months that the main way people are using AI is
Voice and so you know as a as a business leader executive
Employee you should be thinking about how are you gonna be able to use these tools in a year from now with voice that you can’t currently with
Text and and how is your how is your life gonna look differently when you can just talk to the AI and have it help do work for you?
You know, you’ll even be able to do things like you know
Create AI agents where you’ll be able to send off the agents to do a little task for you and command them by voice
Right, so like that’s coming very very soon
So like imagine, you know be planning for that and you’ll be in a way better position
Than people who have no idea this technology exists and if you have a business that has content online
It’s really easy these days to make an audio version of that written content as well
So I think a lot more people are going to also consume content via audio, right?
I think the prompting is gonna be more audio-based where we talked to Siri
We talked to Alexa we talked to these tools and it sort of goes and does the prompt based on what we say to it
But I also think the reverse is true where over time more and more people might consume their content that way as well
So if you have a blog with written content
Throw that content into 11 labs and to have it a podcast audio version that people can listen to as well
Because now you just have another format that makes it more likely that somebody’s gonna consume the content that you just created
So I think that’s another like take away for businesses listening to this is lean into this use this technology
It’s actually a really cool way to make audio versions of your content
Yeah, and I think you know in terms of you know people should be playing too for like how they can use this internationally, right?
Like this is gonna open up so many opportunities that your business can’t currently take advantage of
You know think about like okay people who speak different languages
I’ll know be able to have business meetings with them or think about if you’re making videos or written content that you didn’t turn into audio
You’ll be able to turn that into like a hundred different languages
Right like what does that mean for your business?
So I got so everyone should be thinking about that right now and hopefully we’ll do the same with the podcast
Hopefully we’ll have this in Japanese the next six months or something like that brings up another question when you proposed
Did you actually propose with the cell phone like a translator?
I tried to not use it and then it was it was it was necessary within 30 seconds
But I tried my best and another thing is there the other thing I didn’t mention is you know my son
He’s bilingual. He speaks English in Japanese both perfectly early or at least for a 10 year
You know a 10 year old and and so occasionally, you know
He helps translate which I try to make sure he doesn’t do that too often. It’s like kind of you know
Annoying for him, but but but he but he was there as well like in another room. So I was like, well worst-case scenario
I’ll be like, you know, no
Very cool. Well AI is is changing lives and building relationships. Yeah
Exciting times we’re in bringing loved everyone
And yeah, okay go watch the movie her I’ll go watch the movie her and on that note
I think we can wrap this one up. Awesome. Well, thank you so much to everybody for tuning in
Please like this video and subscribe to our channel if you haven’t already it really really helps get our podcast in front
Of more people if there’s somebody that you know that this episode can be helpful for send them the link
Let them let them tune into this episode if you’re on a podcast player like Spotify or Apple
Give us a subscribe and maybe even leave us a review. We really really appreciate it
It helps spread the word of this podcast and shares this information with more people. So thanks again for tuning in
You
You
You
You
(chiming music)
[BLANK_AUDIO]
Episode 3: Are you ready for speech A.I.? Because it’s here, not in 6 months or a year, but now. AI-generated voice technologies will have a major impact on many things in our daily lives. This technology will effect everything spanning applications in language learning, music creation, emotional AI interfaces, and the evolving landscape of personal digital interactions.
These tools out there now that you can communicate with directly by voice, and there are things you need to be ready for. Matt Wolfe (https://x.com/mreflow) and Nathan Lands (https://twitter.com/NathanLands) dive deep into both the incredible potential and the inherent risks of this groundbreaking tech and how you can take advantage of it.
Check out The Next Wave YouTube Channel if you want to see Matt and Nathan on screen: https://link.chtbl.com/4FZET15d
—
Show Notes:
- (00:00) AI shift towards voice interaction, Hume’s empathetic Eevee impressive.
- (03:41) AI speculation for Siri; potential Apple-Google partnership.
- (07:41) AI technology making it easier to deceive.
- (09:52) Regulation provides bumpers for company behavior.
- (13:12) Cryptocurrency regulations are uncertain, causing concerns for builders. AI regulations add complexities and potential drawbacks. However, voice AI offers valuable content creation tools.
- (18:30) Learning language through translation enhances relationships and communication.
- (21:52) AI music making advances, creating good songs.
- (22:22) AI music tech advances creating music industry division.
- (25:43) Spotify’s payment model raises complexities and concerns.
- (30:10) Immersing in AI detection, improving AI audio discernment.
- (31:57) Content also available in audio format.
—
Mentions:
- Hume’s empathetic voice interface “EVI”: https://tinyurl.com/4cpx2wuw
- Apple’s ReALM announcement: https://tinyurl.com/4uy5suth
- Anthropics $4 billion investment: https://tinyurl.com/3a38cvvw
- OpenAI’s voice replication model: https://tinyurl.com/ahzfkc9n
- Timekettle translation earpieces: https://tinyurl.com/bdv98r39
- Suno version three AI music: https://suno.com/
- OpenAI’s realistic AI voice technology: https://tinyurl.com/4d8mybf6
- Eleven Labs AI Voice Generator: https://elevenlabs.io/
- Her: https://tinyurl.com/y8zkzb5k
—
Check Out Matt’s Stuff:
• Future Tools – https://futuretools.beehiiv.com/
• Blog – https://www.mattwolfe.com/
• YouTube- https://www.youtube.com/@mreflow
—
Check Out Nathan’s Stuff:
- Newsletter: https://news.lore.com/
- Blog – https://lore.com/
The Next Wave is a HubSpot Original Podcast // Brought to you by The HubSpot Podcast Network // Production by Darren Clarke // Editing by Ezra Bakker Trupiano
