Ruby Tso

Something that trains your ear. Helps you catch what you couldn't hear before. The person who knows which part to change, why it needs to change, and then actually edits it, is still the artist. AI can take care of everything else.

What if AI could hear the gap between your song and your reference, and show you exactly where it is?

You don't get better at music by watching tutorials. You get better when someone listens to your song and tells you what's wrong with it.

@stephenm68 Food labels tell you what's in it. For music, that's how much human editing went into it.

Re AI & recorded music: surely labelling is key. Governments around the world ought to be able to legislate to require recordings to carry the details of how they are made, including whether AI was involved. Then the consumer has a transparent choice. Same as with food. #r4today

@thewaleadeyemi New artists now have to ship music consistently, build a brand, grow a community, and stay visible to fans. All before anyone invests in them.

Ever since there are no gatekeepers anymore in the industry and labels can’t engineer an Artiste into stardom like before then nobody wants to gamble with an unknown talents with their money. The target now is whoever has a growing community and the algorithm favors.

@Onilumilulu People will sit through a 4-5 minute song if the first 10-30 seconds give them a reason to stay. The real work is the opening.

Attention span changed music structure. People don’t really have patience for 4–5 minute songs anymore. If you can’t say it in 2 minutes, most listeners are already gone.

Writing, editing, shaping the work yourself, with AI helping along the way. That's still making something. That counts.

@ramnotsheep @PopBase Exactly

@RRRRabbitX @PopBase There’s a huge difference between “I asked Claude how I can make this guitar recording sound more fuzzy” and “yeah a computer did 99% of the work”

Jack Antonoff slams people who make AI music as “godless wh*res”: “to everyone who is gassed up about the new ways you can fake making art, by all means drive right off that cliff. we're genuinely happy to see you go. generations coming will be engaging in the ancient ritual of writing, recording and performing as it comes to us from god. so as we embark on this strange detour where the bad actors will willingly reveal themselves through slop, and the struggling greats will be further spread thin to make an honest living doing what they were put on earth to do, we (myself, the band and frankly everyone i know) remain more dedicated than ever to reveal what comes from within. writing music, recording and performing it - that's it. nothing more embarrassing than considering there is a way to optimize that holy process.”

@Variety The question isn't whether to use AI. It's whether the human is still writing, editing, shaping. If the human led the process and AI assisted, is that not still real music?

Jack Antonoff slammed people who use AI to make art and music as "godless whores": "So to everyone who is gassed up about the new ways you can fake making art, by all means drive right off that cliff. We're genuinely happy to see you go. Generations coming will be engaging in the ancient ritual of writing, recording and performing as it comes to us from God." variety.com/2026/music/new…

@iamfr_lovell Agree. It is built song by song. There's no shortcut.

Most artists quit right before their breakthrough. I preach patience while working aggressively. One viral song can change everything, but it usually comes after months of “invisible” work.

@korymath This assumes the audience stays the same. When AI can generate "good" every time, people might start craving exactly what AI can't do. Stranger, riskier, more personal. Supply changes demand.

As AI gets better, your creative judgement is loses its value. Here's my case against taste...

Music popularity follows the medium. People find new music through platforms and social media. Platforms push tracks. A song that performs gets recommended to more listeners. So the song has to stand on its own. Not the album.

@modhunna Just be yourself. Pretending to be someone you're not is exhausting, and people can tell.

I hate that u can’t just blow up off music n music only goddamn , u need a aesthetic, u need to do some wild shit , u need to do extra shit or else nb gon really care

Your music shouldn't train someone else's AI. Letron uses only clean training data. Every edit you make gets recorded, so your work can be streamed, licensed, and defended. I'm building the AI DAW you can actually trust.

@gina_scooter Part of it is just how distribution works now. Platforms recommend songs, not albums. If a track performs, it gets pushed to more listeners.

You can't tell me music standards haven't significantly declined over the years and then also argue modern albums aren't meant to be cohesive pieces of art but instead choose your own adventure projects where you're expected to discard 70% of the project.

@Jbm_dev The ones that break through will be AI tools that make people feel more capable, not more replaceable.

just overheard someone say: “ever since AI has been getting more popular, it’s getting in my computer and messing everything up” the general public is not as excited as we all are, most are afraid. it’s a barrier to adoption, and we need to fix it, but how?

@Blaqsoundz__ Melody is what people feel before they understand.

A good melody can carry bad lyrics… but good lyrics can’t save a bad melody.

@musicben_eth If the training data isn't licensed, nothing built on top of it is clean either.

Suno is now a $5 billion company and still has almost no license agreements with the music industry, labels or artists.

The AI music tool I'm waiting for lets you actually edit after it generates. The real skill isn't prompting. It's knowing what to change, what to keep, what to shape. That judgment only comes from artists who actually know the music.