Charlie I haven’t been scared in a while since nothing is really that’s scary anymore and unpredictable but oh boy” AI spooky spooky lunge” you got me 😭😂
@Chris Mauck genuine question: what in the world are you talking about 😂 like I'm laughing but I'm genuinely confused, I can name a couple criticisms for this comment but I'm completely lost as to where you found hypocrisy in it lmao
It's genuinely amazing and kinda hilarious just how spooked I got by him literally just moving towards the camera with his hands out and not even shouting, just talking slightly louder than normal 😂 it was just so unexpected
The literal only good thing I can POSSIBLY see from this is for people who have lost their voice being able to have it again. If they want to, I mean. I lost my ability to draw due to nerve damage, and if you told me I could continue to draw using my same style, I'd probably get excited at first, then depressed as I realized I was not able to improve. I'd be forever trapped at the same skill level, unable to experiment, work with new styles, new genres.
You’re thinking about it too much as a replacement for a singer, instead of more like a separate instrument. Think of it like a more sophisticated version of Vocal Chopping, which has given us some absolutely banging music on ITvid in the last 10 years, like my favorite example, Young Scrolls, who just takes voice lines from Elder Scrolls games and just chops them into lyrics, and auto tunes them And it’s amazing, think of what you could do with something like, but more accessible and easier to learn My point is people should realize that you can do things with this you just can’t with the human voice, and so having the option is a good thing
But that’s what’s so scary about most of this ai is that it’s ever adapting and learns from you and creates totally unique pieces of work there is AI that will draw for you with only descriptive words
@Owen Pietromica the point is, is that it will eventually hit that point. AGI is basically an ai that learns and can further its code. it’s not a reality yet, but AGI is like the biggest thing the majority of the ai industry is trying to go for right now. if it improves for too long we hit the technological singularity and we either; live in paradise, or get fucked over by our creation
People are gonna need to put in their wills whether they want to "be immortal" or if you're not allowed to use their name, likeness, or voice via AI. And they will need to appoint someone to sue on their behalf
I don't usually comment on things but seriously dude your opinion on releasing "new tracks" using AI generated voices of deceased musicians really hits the nail on the head here. We'll never be able to trust "unreleased" tracks again.
Especially Juice Wrld bro, now I can't trust "unreleased" music. It's very upsetting and alarming because the person replacing deceased artists voices and making money off of it which that's ridiculously bullshit.
i'm not afraid of AI, i'm afraid of the industries and corporations and politicians that will not even hesitate to use something like this for their own selfish and power hungry gain. it's not the AI that's scary, it's knowing that it is inevitably going to fall into corrupt hands and we are all hopeless to stop it
Probably too late for that if the general public has access to it now theres no doubt they have been using it for much longer, im worried about stupid people and conspiracy theory types that already cant tell the difference between real and fake images online already lol
As an audio engineer, it's a bit concerning for my job with these advancing so fast but I imagine there will always be a place for live music and real artists.
@Juel Anthony well I mean Audio engineers have to worry about audio engineering making money, I mean Jesus man you literally just went “sucks your gonna lose your job but at least I’ll keep mine”
I mean engineers still have to work on the stuff, the AI is literally *just* synthesizing peoples voices, it doesn’t do any of the EQ-ing or mixing with the rest of the tracks, I can’t imagine it’d hit your job *too hard*. Or at least I hope not, I always love working with audio engineers, most underrated part of the music industry man
My issue is more in thinking about the impacts this could have on a lot of legal things imagine getting interrogated but they use an AI like that to make a false confession in your name scary stuff
Yeah but after like 1 case like that is shown to the public there will be laws around that for sure. Then enyone will be able to say about anything "it's not me it was actually AI" even if it was actually you.
@kk they’d honestly probably just end up having to make all evidence provided by the person in question required to be provided on court house grounds which would impede a lot of processes and make actually finding the culprit harder in some senses as well tho
Exactly, now everyone finally agrees that Andrew Tates “phone confession” was indeed fake. We’ve been saying this for awhile now. It’s obvious you can tell it isn’t even him but nobody wanted to speak about it until now. Better late then never though 🤷🏼♀️
What’s scarier is how you can probably combine ai with holograms and it’ll be like you actually revived deceased artists from the dead. And what’s even scarier than that is that any artist of our time could die at any moment, but we’d never know because they could keep it from the public and use ai and holograms to continue their career as if nothing even happened.
The music industry already exploits artists. Now here comes the "oh we have lyrics written by so and so deceased artist so we used AI to create his vision ECT"
Its not just music industries, honesyly its also the animation industry. As AI progresses and becomes better, it might companies might start using it even in the movie industry or so. I follow many artists on twitter and as an artist myself, that is concerning. Yes its amazing, but that means a lot of ppl will be out of jobs just cause companies can just use ai to create things like that
Wouldn't they still need people to refine the art and provide reference images for the art? AI is only as good as it's reference data sets. Hence, I feel like it will act like an autofill feature for artists. Besides companies can't keep churning out the same kind of images before the customers get bored, which is where artists come in. They provide their knowledge and imagination which will either beat the company or gain a high value job at the company which is a win-win scenario for the artists.
@Ah Shit Here We Go Again you can say that now, but what about in the future when ai will be advanced enough? And i remember a while back i seen a video floating around of t2o dudes just acting a few scenes out and the ai maxe it all anime like. Anx while yes it didnt look the best, i can only guess that ai will improve in the future.
@Immortal Revenant That's just rotoscoping, something that's existed since the mid-20th century, they used "AI" to trace their acting and put "animation" over top.
I think voice AI is evil and it’s going to get worse with spam calls all across the US. I told my mother that if you get a call that has the same area code number or an unknown number, don’t answer. These sinister scammers could duplicate your voice and use it against you with voice AI technology. When it comes to the music industry, I could see concerts becoming more popular because if voice AI gets really bad, musicians would rather have fans come out and see them live.
I remember when boomers would say vocaloids would ruin the music industry but vocaloids were just an instrument that showed a ton of creativity. But damn ai may actually hurt the industry
I'm an architecture student and I had a midjourney trial for my studio class this semester. It only does the heavy lifting and dirty work like composing the general atmosphere and arrangement of what you want with very limited precision on fine details, even then it is touch and go. You can type all you want and it might not even give you what you want, even if it is a simple orientation issue that is easily described and understood by humans.
This whole situation makes me worry. Just a week ago, my relative died, who played in the same band with a fairly well-known rock musician in Russia. I wanted him to teach me and my friend how to play the guitar, so that later we could become musicians ourselves. And now every day I see more and more depressing news with this damn AI. I'm not a fan of saying that it used to be better, but some things in a human life should not change.
Please do not interpret my thumbs up comment as anything offensive it's nearly 2am and my brain is completely fried from binge watching all of moist critikal's videos from 6 months ago to now
another scary thing about ai voices if it’s not regulated, aside from the music industry or voice actors. is if it eventually becomes easy and accessible enough that it can be used by anyone on ANYONE with a few clips of just them talking. imagine school bullying, a heavy social media user who just talks online via snapchat/instagram stories or maybe a tiktok livestream then someone able to make a voice mock-up of them saying vulgar things or slurs especially if paired with deepfakes it could be extremely hard to disprove maybe not in a court case as i’d imagine there’s ways to sort that out, but to convince the rest of the school/small community or even family that it was all ai wouldn’t be easy then of course the facebook elderly believing everything they see even if it is a world leader talking about overwatch and such
It sucks to realize we are on the tail-end of a golden age of art, where each and every piece of practically every single artists work was nothing but their own personal imagination (except in rare cases of plagiarism). Even the art in a random advertisement seen on the side of the road was a tiny window into some stranger's life. It's a shame how the human part of art has seemingly been deemed unnecessary these days.
@Big K Spicy As someone who literally spends their entire life trying to make music and work on my music chops it kinda sucks that you think just because some pop music is gonna be sung by a computer that it invalidates the hard work and soul that millions of local and DIY artists put into their work but whatever, keep thinking that everyone is a soulless cash grab I guess
@Owen Pietromica I don't think that it invalidates "real" music. I know that the people who truly care will continue making things like always. What I meant is that we lived in a golden age where even the most crappy low effort art gave us a tiny morsel of human connection, and that golden age will probably end now. "Real" art will still be real, but those who don't care will abandon art for efficiency and just automatically generate something instead. Don't get it twisted, I really value what you do, I'm only bummed that it's becoming increasingly rare.
@Big K Spicy I don’t think it is becoming increasingly rare, we live in a point where more kids want to grow up to be musicians than ever, I mean why do you think the stereotype of SoundCloud Rappers exists? There are millions and millions of people who make music that is never seen or heard by anyone, and they do it mostly for themselves, at the end of the day musicians aren’t getting increasingly rare, you just aren’t seeing them
I feel like the more industries AI takes over, the more affected people there will be, and it will eventually lead to a riot/uprising to stop ai from taking over jobs. Which i personally think is definitely needed right now, seeing as how many jobs are already being slowly taken over by AI, especially art.
I’m thinking of ways to combat this & the most obvious is to put a heavy emphasis on live performances. This could be bad for business in my opinion. People value authenticity over everything. People are our #1 asset. But where there’s smoke, there’s fire. Time to think about ways to fight back against this. It could destroy the music industry.
As a musician since age 10, and as a music producer who has had two years experience with audio engineering focusing on making Excision and Skrillex style bass music, I have to say the amount of work and dedication it takes to be able to actually produce a professional sounding track no matter what genre, or what role you play in the production, is so great that it is what makes the music so beautiful. AI has no emotion, Music is emotion, AI generated music is cool sure, but its stale due to 100% lack of emotion being put into the piece of music. In the music production world, we could talk about how easy it is to make a trap beat but that's a deep rabbit hole not worth explaining here. Selling quality art made by AI and putting your name on it is cheap! This goes for all creative entertainment based hobbies and especially film! Imagine watching an amazing movie with great dialogue and plotting with all of your favorite actors in it and then later read on the CD case "AI generated film"...
Took a class on AI last year about the history, literature, reality, and potential future of AI and it’s absolutely fascinating seeing just how quickly and accurately things have been advancing
YES you literally said what was on my mind before you said it. I am fully convinced that the new Linkin Park track is AI & that all recent releases of deceased artists are AI.
I was working at a game company for a while as a graphic designer and half of the signs we made for open world city games for buildings etc were ai generated, some tweaked by us. There was just not enough time to pump out 400 some building signs never seen before by scratch while also working on stuff like the game package, ui, etc. So yea, I see ai taking over and one day potentially taking a lot of jobs away from people in the entertainment field. Pretty scary
If you watch an anime named “Carole & Tuesday,” one of the main overarching stories included in the world building is that they live on Mars and A.I. is the main thing creating music. Such an interesting take on it though, and this allows for the main characters to stand out as being actual singer/songwriters. Interesting how things are moving that way!!
Now, imagine this from future advanced AI's perspectives. "Humans are scary" "I knew an AI, before they changed their code" "Now, I am not saying that we should kill all humans.."
I think this AI voice stuff is a classic case of people being too focused on whether or not we can. Rather than stopping to think if we should. Insert Jurassic Park quote here.
as someone who wanted to be a musician this just sounds to me like ''another job i will never have'' edit: also this is gonna destroy voice actors one day
As a musician, I actually find this to be very interesting. I feel in order to make something like this work, artists who are using other artist's vocals, should ALWAYS GIVE CREDIT to the person whose voice they're using.
I agree, but what happens when I blend 100 different artist's voices together to make a new voice? We better figure out some rules fast, because I don't even know how you would make a legal argument against that.
@Big K Spicy I feel like it wouldn't be much different than what Porter Robinson does with his voice on his music, same with Madeon. We already have Vocoders and Autotune, and other such plugins to change up how vocals sound on their music, like T-Pain. Only difference is T-Pain can actually back it up with a legitimately good singing voice. At the end of the day art is art, and I feel like instead of AI taking over completely, which it may way, waaaaay in the future, like we may not even be alive by then, but before then I believe we'll see a time of harmony with actual musicians and AI, sorta like how Rap and Rock music were popping off together in the 90's (what a time the 90's were)
Now this AI is not smart on its own it's just a blender it's has a lot of pictures database so it's blending the pictures you requested with the style that it's trained on so it can easily give you any style it's not creating new style it just using what it's found not Google
I tried thinking about positives on the music industry regarding this technology, and couldn’t come up with any, but I did come up with a pretty big W for this technology, devices that bring back voice to those who’ve lost their vocal cords due to injuries and what not, through older videos of them talking.
Ngl I really just wanna go back to the days where music was a couple of people in a room with some amps and recording software. No autotune, no synthesisers and no fuckin AI voices. I miss when music was music and I really do love how there are some bands that don’t use crazy software and really stick to the ‘classics’ I suppose, I love it and you can really hear the love and passion in those types of songs. Something the industry is missing these days. I was also going to bring up Linkin Park (before watching the part where Charlie does) and wow, the newest song ‘Lost’ has Chester’s vocals on it which is really good as it’s not AI and just recovered vocals from when they first recorded Lost but never released it. It’s beautiful and everyone loves hearing such a great voice and icon again. I can’t help but imagine if it was AI, the love , the pain and the passion of those vocals and all the love from the fans would be dead. Especially in such a powerful song, it would be criminal. I hope people come to their senses and never consider AI voices for musicians who have passed.
I got jump scared at the beginning. he's not just a master of not being jumpscared, he's also the master of the art of jumpscare which no AI can ever accomplish.
I was reading the comments as usual and saw the comments about it and then I saw it and I actually jumped 💀 I knew it was coming, just didn’t know when 😂
AI is improving faster than it can be regulated and it's genuinely terrifying to me. My art has already ended up in an AI generator and the idea of other people potentially using my work for their own benefit after I'm dead is something I try not to think about.
i’m hoping this changes the music industry drastically but it relies entirely upon the listeners to decide they won’t want label AI music, and up to real artists to independently work hard to keep focus and i feel we could have a real live/local music renissance if we go the right direction
Great video. I've been writing music for close to 20 years and AI has taken the wind out of my sails. Creativity has been cheapened forever. I liked the world before AI.
people have more than likely said this but, i could imagine this ai being used when a big artist passes away. for example, if this was around when X passed, we’d still be getting more music
Mave is a new AI kpop group. There’s more solo AI and groups which is getting pretty scary. Thinking of how hard a person have to train and how many hours they put into creating music and then there’s AI which does it in seconds
Most musicians write and make music for themselves, with an audience being more of a secondary factor, just because a computer can sing good or whatever isn’t gonna stop any singer I know from wanting to sing or keep writing music. I mean shit dude as a drummer I have been dealing with the thought of drum machines replacing me since I started playing But it turns out just creating art for yourself is a lot more fulfilling than worrying how other people are consume other art
I think true artists cannot be replaced by AI. Sure AI can make a drake song, but it could never replicate a whole Kendrick Lamar album that would come off as personal or as emotional as his albums usually do.
There are already AI generated Chester Bennington songs, and it feels so weird listening to them, just uncanny and creepy, but it sounds good at the same time.
How do I know this isn't an AI created script? But this does bring up a good point that eventually AI are going to be able to make whole videos and audios. It is going to be difficult to distinguish between what is human and what is machine made. I agree Charlie.
At the start when he shouted "AI!" I jumped 3 meters back doing an unsuccessful backflip, causing severe brain trauma and ending up with my full body paralysed. Besides that AI is advancing super quick.
He was so good- I almost jumped off my seat, not really scary but the way he screamed AI while adding those jazz hands made me feel threatened and horrified
Wouldn't really hold up considering how many people have voices basically indistinguishable from each over. Especially if you get 2 artists with basically the same voice. The idea that one person gets the rights to a certain sounding voice just because they are more famous sounds pretty horrifying tbh. The only thing you could realistically trademark is the name. Considering how much hype goes to specific brand names just because they are already popular, we would likely reach a point where an artists name is treated the same. Just as the difference between a Louis Vuitton bag and a cheap knockoff. There is no objective difference in quality, but one already has prestige and as such people are willing to spend more on it. The same with music, an artists sound might be no better than something created by an AI, but simply having the artist involved would rake in more views/money due to their pre-existing popularity.
There is a lot of governance work being done behind the scenes. Artists will be able to "trademark" their voice so it can't be abused this way. Of course it'll take us some time to get there and adjust like how we had to adjust to the "internet" but change is good, for the most part :)
Digital artist here. I'm 100% in favor of AI development, it's all about sharing. In your example of making new songs from dead artists, it's totally fine for me as long as people share it and don't make a profit from it. We are at portas of a great discovery here, saying it's a "bad" thing makes no sense, it would be like criticizing the creation of iphones because it would take even more of our time from interacting with real people and that is bad. We have to improve what we do, and AI seems to allow just that.
I want to see an artist take their voice model, switch it to the opposite sex, and do a duet with it. Like the human is still writing the lyrics and dueting with their "opposite"
This makes me pretty glad I'm not famous... Also I really dislike using dead actors/artists art/voice/whatever through AI... I lost a fairly well known local artist friend last year and ever since everyone says they know exactly what he wanted, (usually something that benefits them) even if it directly conflicts with things I heard him say or do. They're now creating a 'charity' and a venue in his name when he told me he didnt want any recognition after death, he just wanted to disappear. And he didnt care about a venue but wanted a record label that would help others. These people did spend years with him same as me, but they still didnt know that? Or they're just using his name for thier own projects... Just think it's a good example of once someone is gone there's no way of truly knowing what they wanted, so they should be left with whatever they did and no more out of respect.
As a Voice Actor I am scared for my Career Not only because of AI, but because of the Master Class Jumpscare that Charlie showcased, I could never pull that sheer amount of scary off
Better make it yours, before it takes you down, coming from someone who had their job token by AI over a decade ago, learn to use it, and start offering the service and make money from it, now. You got the ear for what's good, you are still at advantage.
Man at this point we're going to get to the point where people are going to have to copyright their own voices. I'm actually for Ai stuff like this, though I think about it differently. If we can replicate voices maybe we can make our own voices eventually. And I can just imagine people who are to shy are hate their own voices using Ai voice technology to preform where they couldn't. Though once more lots of bad things could happen from Ai. But, I still think AI can be a great thing for humanity.
Ok so even if you don't see if personally Charlie, I wanna say this REALLLLLY plays into the idea that Caroline & Tuesday references where one character in the story pairs up with the maker of a powerful vocal AI that basically ends up replacing her.
Crazy how from watching sneako completely destroy himself and what he created in the years previous his tate d riding era now I'm watching Charlie's vids everytime they come out😂
I find it strange how it’s so stigmatised to just take a hard stance against such tech being used in the Music Industry. As you said Charlie, the negative impacts far outweigh any (arguably) positive impacts of AI voice models on the music Industry. So why pretend to be okay with it and say that you don’t want this sort of usage to be banned. It should be! Lol
So, as a game developer/programmer with no skills in art, music or with my voice (and no, not because I'm not "talented" but because I have not dedicated years to practicing those skills and I recognize this.) I think all these AI programs are really exciting. I think it would be great if I could have an AI model trained to fill in all the parts I can't do myself/with my current team. There are also AI models (like GPT) that can write code, and probably write it better than I can or they will be able to soon, and I think it'd be great if a game designer/artist/sound/whatever also had the same kind of tools to help them just fill in the pieces the pieces to help them get from the idea they have to the final product they want to create. I think in an ideal world (and we do not live in an ideal world so pipe dream fantasy ahead!) with this technology everyone would have an AI model that they've trained and continued to train to suit their specific needs and that their AI model will be a part of their marketability/portfolio. Because in reality you could in theory train them to do anything, a basic overview of how most of them work (to my limited understanding) is this: you pass the data or inputs into the AI based on the task you're trying to do (note: I am not talking about training the models, this is what the user puts into it) the data could be anything, pictures, video, audio, text or even data you've already processed into specific types I believe. That data is then taken through what I've often heard described as a "black box" process, the TL;DR is that the AI model figures out the result based on how it was trained and then returns the output it "thinks" is correct to the user. The slightly-longer-but-still-limited version is that it uses a neutral network where the first layer is your inputs, the last layer is your outputs or results (could be just two nodes for true or false or a bunch of nodes like are probably needed to make an audio filter to make you sound like someone else) and the middle nodes are where the training comes in and where some of them diverge into different types I believe but keeping it short there will typically be multiple layers of these middle "black box" nodes that the model will create connections between and at each node it makes a determination and fires off to another node and eventually arrives at an output node (again: this is a simplified exploration and I am not an expert or even particularly well informed on these in my opinion). You train them by feeding them a bunch of data where you already know the results and after it tries to figure it out you "reward" it if it's right and "punish" it if it's wrong and you give it a very large number of a variety of test cases and let it run for hours or days, however the results yield diminishing returns for example the model that beat the world champion at go only trained for 3 hours but it probably wouldn't have gotten much better if it trained for another 10 hours or a few days even (at least compared to how much it costs to train them which is a lot, the cost is going down but it's still prohibitive to the individual I believe) and this stage is usually where the problems and missteps occur. personally I think it's because people are relying on the "magic black box" too much and just throwing it raw input and expecting the computer to come to a logical conclusion. The problem is computers don't do logical conclusions they do *efficient* conclusions, for example I heard about one recently where they tried to train a model to be able to tell a wolf apart from a dog so that a drone with a camera could fly over an area and distinguish wolves from dogs, and after awhile it seemed to be working great except for a picture of (I think) a husky in the snow it kept thinking was a wolf. So one of the people working on it set it up so that it would highlight in green all the pixels the model was looking at to make it's conclusion (because it won't use all the data every time) and it showed pretty much every pixel of the background that did not contain the canine itself was highlighted instead, because the AI found the most efficient way to tell the difference was that typically wolves are outside and dogs are inside and for the majority of it's training data this was apparently true.
I want to hear AI Eddie Van Halen riffs. But Wolfie can release years of real riffs because Eddie left a whole room full at 5150 of discs and tapes of riffs and songs that have never been heard. Tracks that also include Sammy and Dave vocals, besides instrumentals with and without Al and Mikey. Wolfie can even pass it off as his and nobody would know, except for maybe Al. Just a little chat for VH fans.
I feel like the only people that are gonna profit from the labels that already have deep pockets, this is probably gonna make actual musicians and artists completely obsolete
It's has huge implications for just about anything creative. People are using AI to write children's books. Slowly now, but speeding up, people are really questioning the authenticity of modern creative works. This is rather disturbing since artistic creativity is deeply linked to the expression of what truly makes us human and feel alive.
There's going to be gigantic copyright battles in the future for unlawful use of famous people's voices. The music industry will start including ownership of the voice of the artists they sign when they sign them. Things are going to get really interesting really quickly
