AI Voice Cloning Software is Out of Control 😳

AI Voice cloning software is getting extremely good, so good, that it's starting to get dangerous and we need to talk about this. There are PROs and CONs to this kind of tech, and I'd love to hear what you think:
Tools used in this video:
Elevenlabs.io: www.patflynn.com/elevenlabs
Descript: www.patflynn.com/descript
Listen to my podcasts here:
🎤 The Smart Passive Income Podcast: www.smartpassiveincome.com/sh...
🎤 AskPat (these are real life business coaching calls): www.smartpassiveincome.com/sh...
Also, have you thought about starting your own podcast? Check out the #1 podcasting tutorial on KZbin here:
• How to Start a Podcast...
The podcasting equipment I use: www.kit.co/patflynn
My video and live streaming equipment: patflynn.com/live-streaming-e...
Also, follow me at these places below and say hi!
Personal site: www.patflynn.com
Instagram: / patflynn
Twitter: / patflynn

[@AutonomousUltraInstinct69]

we have the same voice, i’m going to file a copyright claim on your channel for using my voice

[@alexanderlyon]

The words "every bite" at 1:30 and "taste" at 1:32 even had what sounded like a microphone pop! That's wild.

[@Innesb]

That’s because the speech is modelled on recordings that had microphone pop. Depending on your point of view, and requirements, the replication of the ‘flaws’ might be a good or a bad thing!

[@Cyber_Kriss]

We live in dangerous times indeed... The potential of these technologies is terrifying.

[@jondeik]

That’s true. I think we need to start an ethics board for this. Because it’s going to become a problem in months. Not years. Months

[@I_hate_mortals]

There have always been dangers with all technology, this is not anything new.

[@GaryStockton]

As a person who loves to do voiceovers, it’s something I feel threatened by. Deep fakes are also a concern.

[@focusedeye]

I'm also a voice talent with similar concerns. I may end up cloning my voice in different style so that if and when I lose my voice due to medical issues, old age, I can still have a revenue stream and after my death, my spouse would still have a revenue stream. Voice cloning is still in its' infancy but is rapidly progressing. I'll give it another while before I take the plunge. All the negative connotations he addressed in this video are the same ones I've been grappling with and have prevented me from jumping on this band wagon. My main concern is where are my "cloned" voice files stored? In the cloud or on my personal computer. If it's in the cloud, what happens to my files if i don't renew the yearly memberships some companies are asking for. Scary times ahead.

[@goingsolointhegarden]

So no problem getting a confession for that crime you didn't commit... This is beyond scary.

[@Cyber_Kriss]

Exactly, it will be used to fool us. Scary times ahead...

[@J3R3MI6]

They’ve had this tech since the 50s and have been using it… same with Sora.

[@isaacsmithjones]

What's interesting is that it'll HAVE to work the other way too. If we're still using the standard of "beyond a reasonable doubt" - video and audio evidence can be plausibly denied. But humans aren't that rational. So if a jury sees it, it may well still sway them.

[@I_hate_mortals]

​@@J3R3MI6Real?!?!?!?!?!?!?111

[@rotorcraft68]

No worries, don't get arrested.

[@TeebeeJ]

You can use this to narrate a book with its different characters using different voices...That would be Awesome!

[@mattredman98]

Great idea

[@heysimmone]

A book of multiple characters being read by one person...nice!

[@davidroyale14]

In my case, I am a latin american creator, and I spend a lot of time doing voiceovers after scripting my gameplay videos, which takes a lot of time for me as my first language is not English. If I can slow down this time by voice cloning, I can spend more time delivering videos to my channel.

[@MikeRussell]

Thanks Pat! This is an incredible topic close to my heart. I hear your cons but was delighted to hear the pros you presented which, I feel, far outweigh any downsides. Another pro that is especially important for creators like us is that you can use your voice clone and translate into different languages. Your content can truly reach a global audience now - all thanks to AI dubbing. That's a feature available right now in ElevenLabs so well worth checking out!

[@80sFootballCards]

I don't know how you can say the upside outweighs the downside. Across the board the downside is by far greater/worse than the upside/positive - political, legal, ethical, pretty much everything.

[@kendalchen]

This is the first video I am watching on the subject (I just found it) and my initial thought is that it is potentially dystopian and terrifying, although as you say in the video, i have also heard that archiving your voice is good in case you lose it in a medical event. I know that in music, fans of some Kpop groups where a member had died were upset when someone used a sample of their voice to record other songs. I create teaching content for my other channel and I like the voice bank at ElevenLabs, but at the same time I would hesitate to make my own voice clone without serious deliberation. This is rough because I think my speaking voice could make some decent passive income, but letting that horse out of the barn? It is a bit scary to ponder unless you could truly hide your identity if you wanted.

[@ChrisDavisTrumpet]

Okay, the intro to this video is ridiculous! Had to watch it twice. When I closed my eyes, I could hear that isn't Pat but if I wasn't paying close attention, I would be fooled. Oh my goodness!

[@Diahadley13]

"I have mixed feelings about this issue. As an industry professional in animation, I'm concerned about the impact it will have. On one hand, I see how voice over technology can help people communicate with others in real-time, even if they speak different languages. However, on the other hand, I'm worried about its misuse and the job losses it might cause for voice-over artists. Big studios that own certain characters could potentially use this technology, which could lead to replacing human voice actors in animated films with AI. This would be a sad day for me, as I think it would take away much of the artistry that goes into animation. Ultimately, I think this issue comes down to whether people prefer humans to do the work or whether they prefer AI. But here's the irony. I've been using Grammarly extensively to fix my spelling and other writing issues. It now has an AI feature that can rewrite things to improve your writing. I'm using it to help me write this post better. :(

[@J3R3MI6]

Voice over artists are done bro.

[@Innesb]

The astonishing thing about this technology is when you record a voice and it plays the speech back using someone else’s voice.

[@thehangingparsiple5692]

Tell that Mr Grammarly his last sentence construct would read more smoothly as: "I'm using it now to better write this post." (I'm sort of teasing here!) But seriously, I personally will never use AI to correct my spelling, or to improve anything I write because it feels like very soon the human brain will become practically redundant. IQ levels have fallen quite dramatically across the world as it is. The time is coming where we will be entirely dependent on technology. I dread the possible chaos and disorder that would follow if this technology failed us. I mean, three supermarket chains in the UK almost completely ceased trading for two days recently, because of errors in card and online payment software. Argh! ❤

[@stevenreid_youtube]

Impressive how tools like 11 labs and Descript are advancing productivity and accessibility for creators. While tools like 11 labs and Descript offer great benefits, the potential for misuse and ethical concerns can definitely not be ignored. Looking forward to further conversations and potential solutions in this space.

[@nancygagnon821]

To have an idea of having the option to hear and feel someone who died, you can watch Black Mirror S02 E01 Be right back ! Very interesting !

[@tesladeathray8198]

mate, that was a possibility of this nightmare on the future. bring someone who was died into this existence.

[@thehangingparsiple5692]

OpenAI have unveiled one that takes just 15 seconds to replicate your voice. It's being kept 'under wraps' from the general public (!?!) until such a time as voice verification is phased out for things like banking. Also, they're looking at guidelines which make it impossible to replicate voices similar to celebrities or political leaders. I can't see how that would work, or even help prevent misuse...in fact, I don't see the point of it at all

[@SovereignKnight74]

Voice preservation! Voices like Frank Walker, and his iconic voice for G1 Megatron and Peter Cullen for the voice of Optimus Prime! Not to mention William Daniel's voice, for K.I.T.T.!

[@TomTheEnglishPicker]

For me the worry is it’s like the dna services ,if you upload your voice, what’s the data protection like for them .

[@Innesb]

Reputable services such as 11 Labs have checks in place; before you can clone your voice, you are required to read a random phrase provided to you, and this is analysed to ensure that it matches the voice you are cloning (i.e. it must be your own voice). However, as with all other AI generators, once there are algorithms in the public domain, there are no protections. I did an experiment; one of my financial apps requires voice verification to complete some transactions, and it will not verify a transaction with my cloned voice. I am unable to determine the difference between a replayed recording of my real voice and my 11 Labs cloned voice when short phrases are played to me, but the voice verification is currently able to reject the cloned voice. When I listen to longer excerpts of my cloned voice, I can always determine that it is the clone and not a straight recording due to very subtle intonations that in some of the cloned speech; intonations that I don’t use. However, the quality of the clone is quite extraordinary, and I suspect that most people wouldn’t be able to determine the difference. The most recent update to 11 Labs algorithm has improved the cloning even further; it’s astonishing. I haven’t tested my cloned voice on my wife, but I will. She is very perceptive, so the outcome of the test will be interesting.

[@bastianrivero]

There will come a day, i will not consume any media whatsoever 😂

[@jillymonson]

From the Desk of Amy Suto showcased this stuff a couple months ago. Her audio book, podcast, etc - none of it is really her, but it sounds JUST like her!

[@heysimmone]

What a time saver for her!

[@nancygagnon821]

Very good for work stuff... So dangerous for what it can do for kids and bad intention people !!! I am so mixed up !

[@AJsGreenTopics]

You are right about having a code word. The world is getting crazier each day. 🤪

[@mattyw274]

Until the computers can figure out that codeword !! Haha Hey. They can do some crazy shit

[@BlastedMaster]

If you use another person's voice like this and said person isn't aware or consenting, at the very least there should be a mandatory display of a disclaimer (or spoken disclaimer if it's audio-only content) proclaiming all dialogue which is ai generated to be as such.

[@tdot8398]

The first thing I'm going to do is create my own sith holochron like Darth revan telling my grand kids all the bs their parents did as teenagers so they can talk back with receipts. 😈

[@koolaidman100x]

I was waiting for it to get better so I use my voice but never could copy it

[@ErikThureson]

There will be both good actors and bad actors as with anything. The tools themselves are morally ambiguous. A knife in the hands of a sushi chef delivers delicious fish for all. That same knife in the hands of a burglar is meant to deliver fear and harm. The knife doesn't get to chose.

[@Just_Samson]

4:06 😂😂😂

[@rossclark5653]

First obvious use case for me that comes to mind is a show like The Simpsons. I feel like the should pay the voice actors as they do now, but use their voices from 20+ years ago. The new episodes the characters sound weird if you haven't watched in a long time. The voices have aged considerably, but the character on TV hasn't. They could continue to reward their actors estate even past their own life if they continue to make episodes.

[@HarryJoiner]

3:04 - smart

[@_ENOVA]

You can make an audiobook or visual novella with voices of the characters, while also localized to your language. Basically - for localization of cartoons, movies, preserving voices of the original actors. That would be huge and very accesible, bringing people of different nations together. letting them share same audio-visual experience.

[@thehangingparsiple5692]

That surely would be a minefield of things like copyright violation, royalties violation, etc etc? Why not just do it yourself as a narrator? This is the art. I'd personally prefer to hear a 'real' recording. Plus how would you get the AI to act certain roles? I'm imagining that would take quite a bit of work ❤

[@thehangingparsiple5692]

I mean 'act certain roles' as in act happy, sad, angry. Wouldn't it be a lot of work?

[@_ENOVA]

@@thehangingparsiple5692 u teach AI with a voice u need, then u record ur speech, then u tell AI to transform ur speech using that voice. Done. 2nd option: u train AI good with samples that contain emotional speech. Good AI voice clonner will do very VERY nice emotonal speech in result. I prefer 1st option cause it give me better control over pace, e.t.c.

[@_ENOVA]

@@thehangingparsiple5692 but yeah - this is really a minefield. But me personally want to try to do next: take an old (favourite) cartoon which was never been translated into my language, use AI to make semi-professional voice over, including women voices. And all done by myself alone. My own voice is no good to listen. These are the reasons why I want to do it. I want to do it for a community of fans (of that cartoon) in my country. No profit. No money involved.

[@mwdrum]

Recent CEO of Open Ai , really non-transparent about this. Basic data mining for years, “borrowing “ your image, videos etc….the casual heft of intellectual property and copyright infringement is a no brainer. If this is your first post about this you’re already behind, it’s moving so fast. Sora, ripx, fadr, Suna,

[@ci3008]

Code words, great idea!

[@agrid]

I am guessing is about who owns a specific sound pattern, not your voice per say. So for example you own your own body, but what if I can produce a clapping just as the one you can produce? I can use the one I produce even if it sounds just like yours, right? In the case of AI, if it is trained with a specific sound, it seems to me that who ever owns that "recorded" sound or its reproduction, maybe the one who gets royalties for AI cloning. But I guess theres a lot of other things to consider. For example, AI uses a way to produce X from noise. So maybe your voice was used to train that AI, but when you train in a Gym you don't have to pay royalties when using a machine to the original creator, right? I mean, I don't know, this stuff seems to be kind of complicated.

[@GideonShalwick]

Yeah it ain't going away - that is for sure. So we are only left with how we can best work WITH it! I think we may need to be super flexible with our expectations here... while sticking firmly to our core (human) principles. I love the idea of "augmented AI", where we use AI to augment and improve what we are already doing as humans. And I am not quite at the point where I want to watch videos completely made by AI lol! Not sure I will ever get there... unless perhaps we get proof that AI becomes conscious and has a soul! But then again... it might be hard to prove that for humans even!

[@SteveTalksBooks]

Wow that's insane

[@TheTenerifeVibe]

it was the helium version

[@rajeevgrover26]

The quality of cloned voice is not good.

[@chanastudley3P]

Can I use this to turn my novel into an audiobook?

[@DeclanKirk]

Such an interesting video! In the last couple of months the software tools to create these types of tools are becoming more and more accessible and easy to use. It took me less than 48 hours to build an AI appointment setting chatbot that can qualify leads for my business- which is INSANE. I'm 100% a blue jedi looking to use the force for good but I'm kinda worried what the bad guys are going to get up to with this stuff.......

[@MichaelSuperbacker]

This is crazy

[@graphicinsights466]

WOW! All very good points. In the not too distant future are you going to believe ANYTHING you didn't actually see someone say?

[@derryXDINES]

Try using the voice to give away some Pokemon cards! I think I shot in the same spot from the same exact angle when I recorded at Genos Steak 😂

[@PapisenPijama]

Bonkers 😮

[@JamesPickering1968]

I am excited about the how it can increase productivity and creativity and though there will be misuse, we'll get control over it. Granted, you will always have misuse out there. You mentioned the use of voicing loved ones who have passed. I lost my wife a little over 10 years ago and as I get older, I feel like her voice in my head is disappearing. Oh, how I would love to hear her voice again. From just a few of the video and audio clips that I have, I would love to hear her just say my name one more time.

[@Innesb]

I’m so sorry that the memory of your wife’s voice is becoming lost to you. Some available technology can take short recordings and generate quite accurate cloned speech, although 11 Labs requires several hours’ of recording to produce good results. The thing is, I don’t know if it’s a good thing for the purposes you describe. In principle, it sounds like a wonderful idea, but something doesn’t feel right about it; I wonder if it might create even greater yearning.

[@thehangingparsiple5692]

​@@Innesbagree

[@mwdrum]

11 labs also different languages,

[@bravo1oh1]

I thinks it's silly that it says 'uh' ..

[@parrylengsavat]

Terminator 2 what’s the dogs name?

[@stevevlahos5469]

Both your parents are dead. 😂

[@GroupXIII]

give step-by-step tutorials, we know that you get wood.

[@DashISCool231]

Did You Pay

[@bastianrivero]

😳😳