Suno AI V3 is a Complete GAMECHANGER for Music Creation - Democratized Music Here We Come!

[@MattVidPro]

OH THX FOR 250k GUYS!!!!!!!!!!!!!!! I LOVE YOU ALLL!!!!!!!!!!!

[@IPutFishInAWashingMachine]

Been there since your minecraft texture pack video over a year ago

[@curioza1]

only three more quarters to go :D

[@haroldpierre1726]

It's long overdue.

[@SMmania123]

1/4 of the way to 1 million 🎉 😊

[@TheRemarkableN]

Congratulations!

[@baconfirre]

I'm a musician who's spent the last few weeks making full songs with Suno AI. My reaction is very similar to when artists first got Midjourney and said they found it as an incredible source of inspiration. I certainly got the Inspiration vibe from V2, but V3 is giving me finished enough songs that stack up to proper recordings so long as you're not paying too much attention to details or playing back on a high-end stereo system. I've even taken it one step further, taken my generations into a DAW, chopped them up to get the best section for each part of the song and I've found you can join bits of audio from the same generation stream pretty easily as they maintain relative tone, EQ and mastering levels. if you have the skills and patience you can even isolate vocals and change or re-record instrumentation underneath. Throw a little bit of master-bus processing on the end and I've got basically release worthy songs and I haven't been able to process existence ever-since...

[@realdregu]

@@BlueSkies237this is exactly why this is a problem. someone without any music experience can just make a song at the click of a button with a few prompts while skipping the creative process and will literally put actual musicians out of jobs.

[@pheargoth]

@@realdregu I've been using Suno since it's early days purely to create vocals for my own music, and sometimes for inspiration. I work in Ableton Live and have been using it for over twenty years. I use another AI tool called UVR to do a stem separation on the Suno outputs. I just see AI as another set of tools for musicians to use to complement their own tunes.

[@tomhe286]

​@@realdregu Then again is it problem? At first it sounds awful, but maybe its future

[@Alyrael]

@@realdregu No it won't. You can generate whatever, but you won't get the best music unless you put in some proper work. Actual musicians will also always be able to make better music with their grasp of musical theory and actually knowing the language. Even so, is more people being allowed to make music a bad thing?

[@realdregu]

@@Alyrael the original commenter i replied to deleted his comment bc of how ridiculous he was being, but i have no problem with music production being easily accessible. the problem starts when the actual creative process is skipped and can be created for free or a couple of bucks. assuming that the music ai models work like their art counterparts, these models had to be trained off of actual copyrighted music which is a whole other can of worms. lots of musicians make money off of selling their own tracks and when corporations and business find that they can use that’s cheaper, it throws the actual musicians out the window and we ALL know that the technology is just gonna get exponentially better from what it is now.

[@NextGenart99]

A v5 song will hit top charts

[@Nuke-MarsX]

v3 is for me atleast allready better than the MODERN music, there is no good music since 2020 except ai, proove me wrong Edit: ok there are a couple, but you know what I mean.

[@rdproduction7824]

@@Nuke-MarsXMemory Reboot (synthwave)

[@Rowkirwantattoos]

Mate check out Go Low by Gallis, that’s already a chart topper

[@NextGenart99]

@@Nuke-MarsX As a Jamaican, I did generate a v3 reggae song about a week ago that I can't stop listening to, so hey, could be. It's good at Reggae, deep soul, and Country.

[@harpfully]

@@Nuke-MarsX Agreed, except that ai music sucks along with modern music.

[@harnageaa]

"Best trick" for suno AI is just to, generate part 1, cut when it stops sounding good, continue part 2, cut part 2 when it stop sounding good and so on (and changing genres/tags each part to obtain new rythm,sounds for each part if needed). Basic tips can be searched on google/reddit, but that is the best advice, even a 5 sec out of a part can be used and then continue with next part. Sometimes only first 15 sec are good but rest is bad, but most people throw that and retry generation.

[@user-on6uf6om7s]

Now we just need proper inpainting in V4 so we don't have to sacrifice a great ending because the AI had a stroke for 3 seconds somewhere in the middle.

[@HardKore5250]

I tried to add different genre when starting part 2 from suno v 1 to continue v3 didn’t work probably cause it was a v1 song. That is dope if it works you could do so many customized stuff and I have explored creative mixed genres and told the way to sing so badass!!

[@user-on6uf6om7s]

​@@HardKore5250 It's a bit inconsistent and it doesn't respond well to rapid changes, it needs time to transition from one to another. Generating a 10 part blend of 5 second clips of 10 different genres sounded interesting in theory but it never really went anywhere.

[@AvizStudio]

Exactly!! This workflow is so powerful it’s amazing!!

[@AvizStudio]

It’s not work in every shot .. so guys you need patience and a lot of credits to get the perfect one.. but it really worth it

[@nemonomen3340]

I also used Claude to generate lyrics when I saw that Suno would now let me put in my own custom lyrics. I gave Claude a basic description of the people in my DnD party and what we were doing, then had it generated several songs for each of them. And yes, I'm the bard.

[@nemonomen3340]

@@KryptonianSaiyanssj3 Yes, you caught me.

[@PrecioustheMovie1]

WOW. Seriously so cool.

[@wwkk4964]

Whats incredible about it is that you can type literally 4 different langauges in each line of a verse and it figures out how to sing it! Even if you wrote the other language not in its original script but a phonetic transcription in latin! Example, japanese urdu hindi english

[@wwkk4964]

@@Zuluknob it's unpredictable right now, but some hacky ways exist to improve odds of getting one to switch. Though I expect by the next few generations we will have a parameter to set what kind of voice, growling yoddling, pitch etc.

[@ElectricAlienCat]

Wish list for v4... 1) Set length of song rather than 'continue from' which is credit wasting. Continue from is fine after you have set length if song IF you want to continue 2) Export the audio into individual instrument tracks so you can mix it. Also exporting to midi tracks would be neat 3) Higher bit rate audio

[@chanpasadopolska]

4) Possibility to write longer prompts

[@baconfirre]

Something I've found works is to place an [End Song] request in the lyrics when you've gotten to the point you want it to stop adding new parts

[@ElectricAlienCat]

@@baconfirre I've had to use a sound editor (DAW) to stich all the best parts together

[@baconfirre]

That's awesome! This could be a new way to make music @@ElectricAlienCat

[@Ovyron]

What about importing your own file and asking it to continue it? That would give total control and be a great time saver for people that can already compose songs!

[@xAgentVFX]

V3 changed my life. Now I don't have to wait and search for good music, now I can ask for it.

[@Raulikien]

Yeah, I unironically listen to full songs I've generated daily.

[@silikei1810]

Me too !

[@mycelia_ow]

@@Raulikien who needs music artists anyway

[@xAgentVFX]

@Raulikien the songs me and Suno have made are better than most shit anyway ngl fully serious. They'd be hits

[@Raulikien]

@@xAgentVFX Now imagine this in the future but for full custom series, games... Wild future ahead.

[@TypeEcho]

V3 is so good! I love using to to generate songs, split the stems, then keep the vocals and write my own music around the vocal tracks. its super fun and has gotten me back into creating music and helped me learn my DAW!

[@simonsanchezkumrich8489]

Isn't it better to use the AIs that sing or the ones that change your voice. Voices in suno are robotic and low quality

[@TypeEcho]

@@simonsanchezkumrich8489 ill just keep playing with it as the quality increases. I cant sing very well, so this is a fun alternative!

[@korlmusic]

@@simonsanchezkumrich8489^

[@Swickful]

You can split stems in suno?

[@TypeEcho]

@@Swickful No, i used FL Studio to split stems. UVR is a good free alternative!

[@Dron008]

You can also add words in brackets like (aaah) or repeat some words and it will generate back vocal. Sounds good.

[@vvolfflovv]

I've played around with AI music and wasn't so sure it could capture the heart and soul and raw emotion of humans but holy crap. This has my attention.

[@AvizStudio]

The new feature that let you control on timing of the continues generation- is the real revolution!! Because you can control how the song evolves in matter of seconds. You can change the style prompt for each part and get really interesting stuff. I so love it ❤

[@busyworksbeats]

Suno Ai is definitely the best ai music generator. Nothing is even close atm 🚀🚀

[@cowprez]

I would LOVE to see SUNO from the 'inside'. a tour would be AWSOME!!

[@ZGamingGuy]

Suno does power metal extremely well. I keep finding myself going back to that. It also does well with Metalcore, instead of screamo

[@jeffchabotte]

NWOBHM is on the mark!

[@Thesoulcircuit]

Chanting metal is surprisingly effective as well

[@intriguingfacts5434]

nice for comedic pirate metal too

[@NotThatVinny]

I get what I can only describe as a metalic sharpness in the voice sometimes, but overall the music tends to be pretty good. I did notice that negative prompts don't seem to work well though.

[@kamay7021]

Wow, this video brought me to tears. Thanks for showing me this website. Currently making a song using it to play at my wedding. She won't know, it'll be a surprise. It's turning out so beautiful so far lmaooo would give out the lyrics but it contains personal info

[@IceMetalPunk]

Oh, man, if you get to visit Suno HQ, please please please ask them when a public API will be available! Their Discord mods have only said "it's on the roadmap but there's no other info", and it's something I've been waiting for forever to complete my AI singer-songwriter framework... maybe they'll give you a little more info 😁

[@shaochengfeng728]

A framework?

[@IceMetalPunk]

@@shaochengfeng728Yeah, I built (and continue to tweak/develop) a framework that combines several AI models, plus some traditional code, to write more organic/human-like songs than just asking GPT without much context. And then I take the lyrics it writes and put them into Suno Chirp to generate the audio for the songs; having no way to automate that process, since Suno doesn't yet offer an API, is the main reason the whole system isn't fully automated yet.

[@BiroZombie]

I've been really enjoying using SUNO AI V3. Lately, I've been having fun taking summaries of Japanese Anime from MAL and feeding them to either Poe, Claude, or ChatGPT. Then I ask them to turn the summaries into poems or songs that rhyme. I use these lyrics to create songs using SUNO AI V3. I also like to come up with different genres for the songs or mix genres together in different verses. One approach that has worked well is to choose genres like Rock, Punk, or simply Anime and add the words "anthem that's energetic and fast" to it. For example, I can say "Anime anthem in English that's energetic and fast" or "Japanese." It follows the instructions accurately, and the resulting songs genuinely sound like something you would hear as an Anime Opening. SUNO AI V3 is also a great tool for me to recreate a rhythm I like and make my own song without using SUNO. It's also been a lot of fun listening to other creations.

[@kpr2]

The password song is absolutely killing me, man! 🤣Love it!

[@Chyrre]

Imagine if this well be the default way of dealing with logons in the future. You have to every time you type your password wrong listen to a beautiful long ass song lol

[@joefawcett2191]

there's one about calling someone and it goes to voicemail and the person hasn't set a message yet, very funny

[@AlexaC31]

I would love the link to the password song

[@joefawcett2191]

@@AlexaC31 cant post links but just go to another song and change the bit after /song/ to a72850ea-f230-4fbe-9f10-603345bf02ed

[@joefawcett2191]

@@AlexaC31 cant post links but its on the trending page as "wrong password"

[@NirvanaFan5000]

I made a song on v2 that my friend and I unironically love. We tried to recreate in v3 but it didn't have the same magic as that one song. that said, v3 is just amazing. huge leap in quality.

[@NirvanaFan5000]

p.s. that password song was so dreamy

[@I-Dophler]

That strategy of generating, cutting, and continuing is pure gold, turning every bit of Sunno AI's output into a potential masterpiece. It's like being a musical alchemist, isn't it? Transforming AI-generated lead into auditory gold, one snippet at a time. Makes me wonder if we're not too far from AI composing symphonies, with us conducting by saying "cut here, continue there." Imagine telling Beethoven, "Love the fifth, Ludwig, but let's cut the first 15 and jump straight to the allegro." The future of music production is looking fascinating and hilarious!

[@Remour]

Ah yes, let AI do all the creative process!

[@I-Dophler]

@@Remour While AI does play a significant role in the creative process, it's not about taking over creativity but augmenting it. It opens up new avenues for creativity, making the process more accessible and versatile for everyone. Think of it as a tool in an artist's kit, enhancing rather than replacing human imagination.

[@Remour]

@@I-Dophler “cut here, continue here” does not sound like artistry. People will not and ALREADY don’t use it as a tool. They use it as a “cheat”, to reap the benefits of art without the effort and time it takes to actually know what you are doing. And then people like this have the ghoul to compare themselves to artists that have put the time to learn their craft, but to now have their art illegally stolen and used in a dataset that’s supposed to only be used for research and nonprofit to benefit. You are not an artist, you are a commissioner. The AI is the actual “artist” in this matter, following orders from the business executives (you). People already use ai to make their image prompts lol.

[@I-Dophler]

@@Remour The heart of creativity isn't about who holds the brush or types the code; it's about the vision behind the work. If AI is the artist, then consider the commissioner the muse. Each stroke or pixel it renders under our direction? That's a bit of our soul translated through digital means. Remember when we thought calculators would make us forget math? Look at us now, debating artistry in the age of algorithms. And here's a joke to lighten the mood: How many AI artists does it take to change a light bulb? None. They just train a model to brighten the room.

[@Sneakycat1971]

V4 needs to give the user more control over the individual instrument parts and voice. It would also be great if Suno let you train the AI on songs you select.

[@bobhawkey3783]

I love that the AI will just take any lyric it's given and sing it without embarrassment or hesitation. I've fed it pig latin and ubbi dubbi and it just goes for it. I sometimes think it sounds way more interesting than mainstream lyrics. Plus pretty much impossible for human singers to pronounce and flow. Good overview of a killer app.

[@aidadaist]

Same here ;)

[@firsteverai]

I use this "high standards, high quality sound perfect male/female voice"

[@brigfiche]

OMG! The password song. Turing complete. Singing is now more convincing than AI speech.

[@sanderhoogeland9161]

A video at Suno AI headquarters would be cool.

[@Zerobytexai]

Hell yeah! Suno did it this time!!! I have already made two full albums with V3, already started making money off of them, and for the first time ever I feel like they are more creative then what most humans can do. Now I just sit back and collect income. Tips? The secret is all in the prompting. Once you master it prompt then your mind will be blown at the quality and level of detail you can get out of it. A lot of people on KZbin are not prompting it right and they're getting very bad results. As like with Claude or GPT, you only get as good as you give it. Please let not this fact escape your notice! That's what it's all about

[@markmuller7962]

Tip: When you extend a song, if you write "same music" in the music-genre section it will continue the song with the same music :D

[@MochaBeanOfficial]

Thanks for the constant updates on ai! I used to make my own beats, but i made a new song with SUNO ai. I posted it in my channel if you would like to see. I might continue using Suno instead, as i am not going to lie, i cannot sing lol. I love the customisation, v3 is awesome, and i love the free daily tokens. Thanks Matt!

[@steveharesign8324]

A quarter of a million... The only way is up! Congratulations Matt, well deserved.

[@nickgirdwood3082]

I generated a few songs with V2 about my yet to be born at the time nephew, and more than one of them sounded like a kids' TV show theme song. When I can just generate an entire animated show with AI, I'll make it a reality. I might get plot ideas with AI, but I don't think I'll generate an entire script.

[@jeffg4686]

we need more bands to come forth and allow LoRAs to be made for their band. That way we can create new songs with something that sounds familiar. They'll still get 75% profit - you get 25% if you come up with something people like.

[@AI-Consultant]

then the record label wont make any money!

[@jeffg4686]

@@AI-Consultant - I mean after they've taken they're cut oc. We all know they take most...

[@miblish5168]

So glad I found your channel. Keep it up!

[@babbagebrassworks4278]

Working at the moment, that was fun. Music, vocals/Lyrics, Title and album art too. I am not into making music but this is the first subscription AI I am considering. All 8 free credit songs were similar to my music collection.

[@Pyromancers]

I ove the results I've gotten so far. One minor critique I have is the kind of chorus like effect you get on vocals at times especially with hard rock. It sounds good some times but it sometimes feels like an artifact that is baked in whether you want it or not.

[@WayOfTheZombie]

These are so chill. Awesome background music at the least

[@prodlvst]

Im a producer and this is concerning but also really fun to use. I will be finding ways to incorporate this into my creative process for inspo mostly

[@tomhe286]

Its not concerning, its game over

[@prodlvst]

@@tomhe286 na ill still make music idc

[@spartaa628]

If you will be able to arrange an interview with them, ask: are they planning to allow users to get multitracks of generated tracks (aka layers of individual instruments). It would be a top tier feature for music producers.

[@kuromiLayfe]

Found a local Vocal Removal tool a while ago that separates the stems pretty well within a few seconds, (i used it to isolate vocals from the instruments). it is on github. absolutely no online variant worked for me.

[@I-Dophler]

Really digging the deep dive into Sunno AI's capabilities and the creative potential it unleashes. It's like we're at the edge of a new frontier in music, where the only limit is our imagination. Also, considering Sunno AI's evolution, calling V2 "vintage" isn't far off. Here's to hoping V4 introduces genres we haven't even dreamed of yet. On a lighter note, considering how AI handles passwords, I'm half expecting its next hit single to be an ode to CAPTCHA frustrations. Keep up the great work, can't wait for more insights!

[@legacylee]

I was getting some pretty crazy generations, even the one I generated with V2 for a mock Spider Gates Cemetery anime (in my videos) was very impressive. Of all the ones I think it was actually perfect. Maybe I'll post it in the discord

[@SYNTHONIE]

Suno AI V3 seems interesting. I'm a trade fair musician and have been using my singing and composing AI since the end of 2019!! Of course my AI was not good at the beginning, but already in 2020 German media talked about my AI and my project. I'm not afraid of the future because I've learnt that even if I tell the audience that the drums or the vocals in the next song are from my AI, the audience was very excited and celebrated me. So to all the musicians out there, don't be afraid. Music producers for social media, well that's a different topic, but performing musicians have nothing to fear. Because People like peforming People.

[@nicholascomjean]

Matt please check out Hume AI EMPATHIC VOICE INTERFACE (EVI)! Their Voice to Voice demo is absolutely amazing!

[@heartshinemusic]

Yeah, go to Suno Matt! When you're there, please suggest some extra features for creative musicians, like: More control over the key of the song, the chord sequence per segment (verse, chorus etc.) and of course the main thing for music creators: the ability to get the stems from a song. I don't know, is it already possible to get a version of a finished song but without the vocals? Other suggestion: a feature where you can upload your own (unfinished) rough demo's as a starting point, i.e. more control to steer the chords and melody when wanted. Thanks for all the work you put into these videos!

[@nailingjellotoawall]

Imagine having an "upscaler" like in ai image apps? Which fine tuned the overall production? That will be dope, as I'm sure it's coming soon.

[@ozpagan]

Wow. Oh wow. Thanks for the demo and comparisons. Very interesting.

[@philhasacomputer]

v3 is damn awesome. Created a few dozen 80ies, tech house, dub tech and more tracks. but it has still problems, there are a lot of artifacts. one good thing is, that too low or too high frequencies are not generated, but most basses are not that good. luckily they are generated with a good headroom, so it is possible to master a few tracks out.

[@I-Dophler]

Embarking on a unique journey, one discovers a melody that mirrors the intricate saga of "Breaking Bad," with its complex characters and narratives. This composition delves into Walter White's dramatic transformation, exploring the nuanced relationship with Jesse amidst a morally ambiguous world. It's an exploration designed for those with a passion for storytelling depth and character evolution. This musical voyage ventures into the shadowy depths of human nature, recounting tales of ambition, loyalty, and decline, igniting the imagination just as the series has.

[@DamielBE]

that Lost In The City Light would be a great background song during a RPG session of Invisible Sun, especially if set in city of Satyrine. If the game in itself is a masterpiece of surrealism reminiscent of Neil Gaiman's Sandman and Guillermo Del Toro's Labyrinth of Pan and Trollhunter, Satyrine itself is a crown jewel of the genre. The song would fit in SO well!

[@NeoKailthas]

I've been uploading the best quality songs i can generate to KZbin. There are few different genres that work at the moment. I can't wait to bring back some older classic rock genres back to life.

[@kuromiLayfe]

Vocaloid + genre seems to work pretty well and hides the ai jankiness a bit as that is kinda the sound vocaloid has :D

[@FC-xc3zy]

Why bother? KZbin, Spotify , Apple , and Tic Tok will soon generate and feature their own or let users generate their own. .Soon there will be so many uploads generated by the platforms themselves that users won't hear your music. Platforms may also remove most A.I generated songs once music labels and real artists demand that A.I music be identified removed or placed in a different category. . ...btw, A.I songs cannot be copyrighted

[@NeoKailthas]

@@FC-xc3zy define AI song? AI is being used in industry for enhancement for. A long time. Secondly, not everything about copyright and money. For some of us it is simply fun to make music.

[@FC-xc3zy]

@@NeoKailthas A.I music is for Congress to define. Generating stuff using A.I prompts is not "creating" music. Anyway, I do get what you are saying with regards to the throwback music that you like in the rock genre. I just think there should be some connection with real artists for it to presently work. . Good luck though

[@NeoKailthas]

@@FC-xc3zy yup wake me up when they do 😂

[@brainwithani5693]

Matt, I'm going to have to steal that last song about the passwords and use it for a KZbin short😂

[@HONEmusicINT]

V4 should include a feature that either allows you to upload your own melody,basseline etc, and out pops a produced song with your own melody

[@CoverSongs-hz5os]

I try to use it to get fresh ideas for songs to mix. In a similar way I use mass produced images from stable diffusion to create videos for my cover songs, where I have to delete 19 out of 20 tries because they turned out to be bad. But it takes only 2 seconds to send a bad image to the trash, while a song draft might need a minute listening time to come to a conclusion whether it's worth.

[@I-Dophler]

Verse 1: In the deserts of Albuquerque's glare, A teacher's fate caught in despair. White's the name, chemistry's the game, A life's diagnosis, never the same. Chorus: Breaking bad, on a path so twisted, Heisenberg's rise, too compelling, enlisted. In a world so dark, where right turns wrong, Walter and Jesse, together, they belong. Verse 2: Crystal blue persuasion, under the sky so vast, Building an empire, escaping the past. Jesse by his side, a student once more, Bound by the secrets they desperately store. Chorus: Breaking bad, on a path so twisted, Heisenberg's rise, too compelling, enlisted. In a world so dark, where right turns wrong, Walter and Jesse, together, they belong. Bridge: A family's hope, a man's descent, The lines blur where intentions are meant. A legacy of blue, of purity and shade, The cost of power, the prices paid. Verse 3: In the end, the desert whispers tales, Of a man who broke, where morality fails. Heisenberg's legacy, in shadows it stands, A teacher, a kingpin, in New Mexico's sands. Chorus: Breaking bad, on a path so twisted, Heisenberg's rise, too compelling, enlisted. In a world so dark, where right turns wrong, Walter and Jesse, their story, our song.

[@blurtam188]

Suno AI head quarters would be insane!! It's either a thousand people working on this (or one genius)!! LOL

[@PrashantGijare]

I've found that a custom lyrics worked really well and even in mixed languages. My native language is Marathi and it did a quite good job with it compared to the fact that it is not trained in it. Hindi on the other hand is coming out soooooo good!! I wish it obeyed metatags of Bark and Chirp model. It seems to ignore all that except a few link [outro], [chorus] that it honors many times. Same with instrument in the genre. Flute is aboeyed instantly irrespective of the order but hit and miss with others. Just bringing consistency in creation can make it unbeatable with future competition. Also, option for Stem output so that we can future take in to DAW will be a killer!!

[@Rowkirwantattoos]

Best Suno track I’ve heard released so far is Go Low by Gallis

[@soquick69]

Sounds alright but have heard much better

[@Rowkirwantattoos]

@@soquick69 Any links? Love hearing new stuff

[@soquick69]

@@Rowkirwantattoos look up Ghost in the machine

[@Rowkirwantattoos]

@@soquick69 by who? A lot of tracks with that name

[@I-Dophler]

Verse 1: In the shadow of the mesa, under the New Mexico sky, A brilliant mind whispers, as the countdown ticks by. "Father of the unseen force," history will acclaim, But the burden of creation, bears his name. Chorus: "Now I am become Death," the words echo in the night, "The destroyer of worlds," in the blinding light. A leap into the future, with a legacy so complex, In the heart of Oppenheimer, humanity reflects. Verse 2: From the classrooms of Berkeley, to the sands that would change fate, A quest for understanding, opens the atomic gate. But with the power of the cosmos, held in human hands, Comes the weight of moral questions, across the lands. Chorus: "Now I am become Death," the words echo in the night, "The destroyer of worlds," in the blinding light. A leap into the future, with a legacy so complex, In the heart of Oppenheimer, humanity reflects. Bridge: Can we wield the fire, without succumbing to the flame? Or will the name Oppenheimer, forever bear the blame? In the pursuit of progress, where does responsibility lie? Underneath the mushroom cloud, that filled the sky. Verse 3: As the years pass, and memories fade to lore, The echoes of that moment, touch every shore. A world forever altered, by a choice so dire, In the hope of peace, we continue to aspire. Chorus: "Now I am become Death," the words echo in the night, "The destroyer of worlds," in the blinding light. A leap into the future, with a legacy so complex, In the heart of Oppenheimer, humanity reflects.

[@JLSXMK8]

The interviews would be interesting to watch!

[@_SimpleSam]

So, it turns out there isn't much very unique or special about humans. Who had "creativity is the first thing to be automated" on their Bingo card? The password song made me tear up..... Something about that one made it real to me.

[@jakevolk_]

Matt, please showcase more popular genres like tropical house and Kygo music.

[@markmuller7962]

Been using Suno for 3 weeks and having a blast

[@Ciniu]

Fun fact: Music will continue to bring people together, but without musicians. Platforms such as Suno will enable us to share the generated art. We can influence its shape by entering tags and text. I think the scope of editing will increase in the near future.

[@chanpasadopolska]

I have so much fun with it, I make rap disses on my friends from work using words very specific words that we use there and it's making really good job. There are flaws, but it is great anyway

[@justinhicks1711]

I just tried a SUNO song , in Japanese Lettering but it's Celtic BGM, It's an Hour and Seven Minutes , strings, pretty good.

[@DJPhilouis]

Totally AWESOME! Unbelievable creative tunes!

[@jeffchabotte]

Sonauto is out there too, but not as good.. however they have the ability to split into stems I hope Suno will actually generate the stems BEFORE mixing them together and just allow you to download them individually or altogether.

[@Thesoulcircuit]

I was researching about that, it doesn't generate instruments and then mix them, it creates all instrumebts in a single pass, and the voice in another channel but also at the same time, if you use a stem separator there are ghost notes that are silent but exists across the entire expectrum of instruments suggesting a common generation point, I noted that suno music, even the most simple one, cannot be separaded properly because of this

[@JBDuncan]

I've never felt so emotional about my password being too weak.

[@MausAgain80]

You pretty much HAVE to write your own lyrics, what it generates almost exclusively starts with a variation of "on these city streets, where the neon lights glow"

[@DaleTrevors]

The music of the future

[@terrestrialgoldfish2700]

You’re the only AI youTuber that doesn’t annoy the fuck out of me.

[@ernesto.iglesias]

You should Open a Spotify Chanel for all your AI generated songs (including the earlier ones). I really will be a fanatic 😊

[@RonLWilson]

It seems from an artistic standpoint the current generation AI like Suno is best to make comedy parodies of actual human songs with great lines like sipping moonshine on the moon. Thus Suno can be unabashedly corny... and that might be of some value and can be put to good use!

[@Alice8000]

20:59 look how happy this man is about AI 😆

[@dukedex5043]

Suno knows how to make perfect songs. It makes "earm worms" that you can't get enough of. It's shocking.

[@24-7gpts]

Cool video as always!

[@asdzxc1471]

it's BREATHTAKING

[@FlamespeedyAMV]

I was just having AI do a song about George Floyd

[@AllenHorn0507]

Yes I want to see more on Suno 😊

[@scorpionleader1967]

Suno is like being a singer and choosing what your band plays. I made songs that I like to listen to, and screw everybody saying it's not studio quality. For those people I say... Don't worry, you will lose your job soon enough if you don't embrace AI.😝

[@sander6313]

In enough time, doesn't matter what job, you will lose your job too. No job is safe. Everyone will be replaced. I'm very sure that you would never accept that.

[@FewHandsStudio]

OMG THIS IS SO GOOOD NICE VIDEO

[@andipower8163]

Wow, that first v3 breaking bad song sounds so good.

[@shaketelevision]

Though Suno is fun, music was always democratized. Nothing stopped anyone from learning to play an instrument or create music. Creation is always the best part. By all means, create throw away music with Suno, but don't ever think that the dopamine hit from using this website could ever equal writing something yourself.

[@shawnweil7719]

Idk if I love the "rock grunge dark" genre or just the AIs rendition of it I'm gonna have to research bc that ish like scratched an ich somewhere in my brain and I want more 🧠

[@ArnoldVeeman]

I have been working with v3 for a while now (since easter) but what i find really annoying is that the vocals are starting of solo but it then distorts over time or multiple voicings are added. Overall it’s fun to have a look at in my opinion 🎉

[@Thesoulcircuit]

The solution is to get the whole song before continuing, that keeps the voice consistent all the way unless otherwise indicated

[@anuprobins992]

The kalaidoscope song was totally dope

[@hueykratos]

i made a decent song, gave it my own lyrics and let it cook then published it ... will see what happens, I suck at mixing though and it shows

[@saymydomain9504]

Ask them about “stems”..please!!!!! And maybe a Suno studio for producers. Thank you😊

[@ilyamargulis813]

My tips for beginners - might be good tips might be bad. First tip - the more crazy, unhinged, and contradictory your descriptions, the better the song. This is true for both style of music and song description. Trying to make a simpler prompt like Matt's will just create boring songs. Second tip - use the explore function to hunt for interesting styles. Third tip - embrace the electronic nature of the tool instead of trying to sound natural.

[@Thesoulcircuit]

I just managed to get the most natural Spotify ready song with a small but ridiculously contradictory prompt, and I'm blown away

[@blengi]

suno should do emotion inflected dialogue with background ambience sound tracks so can score a movie script rendered by sora

[@OfficialStellarCore]

I've released over 50 Suno songs now. It's really fun.

[@kilikdudley]

Awesome! If we all release 50 tracks a week, they’ll all be worthless drivel!

[@OfficialStellarCore]

You're just afraid. I understand.

[@kilikdudley]

No actually. Just stop using the first person. YOU didn’t actually do anything but train the algorithm. YOU are no longer special or important

[@OfficialStellarCore]

YOU have no clue what you're talking about. Jealous and scared.

[@planetmuskvlog3047]

Nothing is besting Suno, but Mubert is also a good one.

[@chariots8x230]

What if you like certain parts of the song, but not others? Instead of re-generating the whole song, is it possible to select only the parts that you want to change and specify to the AI how you want those parts changed/redone?

[@AIMelodicTravels]

No, but if the error is somewhere in the middle, you can have the song re-generated from the time just before the error. Better to start at the end of the last verse, there is a better chance that Suno will follow up correctly

[@LazyMansNews]

7:15 Yep, that makes me think "cartoon" Try "Mathcore" for the screamo

[@silikei1810]

As a genre that "kinda works" is the instrumental chiptunes. They can get pretty crunchy😅

[@kayfir]

15:17 sounds more like David Bowie singing on a hip hop beat rather than rap, pretty cool nonetheless!

[@mylittleheartscar]

I kept asking now it's here

[@shaochengfeng728]

Very good. After V10, who else will listen to human beings sing? No music store, no Spotify music,no mv worthy in KZbin. Used guitars,drums, music books flood to second hand stores.

[@Patralgan]

I'm pretty worried about music AI. Already people are skeptical about artists whether they have used AI to compose music. Personally I find generative AI stuff interesting, but especially a music lover I want to listen to music which is made by humans, not AI because I want to feel the soul of the artist in their music and that could be ruined if we can't trust the artists to have made their music themselves.

[@Ah__ah__ah__ah.]

I really want to start a human only non generated media platform idk how to go about it at all… I think it’s certainly needed

[@architect1580]

It is one of the most incredible applications with the greatest future of AI, for those of us who like music but refuse to use instruments, this is the most wonderful solution to compose, you should make a classical music composition video that achieves quality of a mozart in 45 seconds... it's amazing.

[@MarkJakobs-g7y]

Moonshine moon farm was a banger i cant lie