Say you’ve just come back from a weekend at the beach with a phone full of videos. You want to cut them into a short film for friends and family, but something’s missing, and it’s the music.

You could spend a couple of hours going through music libraries looking for a track that fits. Or you could tell an AI assistant what you have in mind, something like: “Make something calm and upbeat, with soft piano and acoustic guitar, about a minute long.”

If the assistant is connected to the right tools, it can help turn that description into an actual piece of music. This is starting to happen now that ChatGPT and similar assistants can work with creative software, mostly through a technology called Model Context Protocol (MCP). For people who have never played an instrument or been inside a recording studio, it could make music a lot more approachable.

What MCP does

Most people know ChatGPT as something that answers questions, writes emails, plans trips or helps brainstorm. Lately, AI assistants have started doing more than talking. Instead of only explaining how to do something, they can connect to compatible apps and help do it.

MCP is what makes those connections possible. It lets an AI assistant communicate with other software, a bit like a common language that different programs can all use.

Normally ChatGPT might explain how to edit a video or arrange a piece of music. With a compatible connection, an AI agent could work inside the actual software and make those changes itself.

The AI isn’t creating everything on its own, though. It helps people run existing tools by describing in plain language what they want, and music software is one of the places where this is starting to work.

What people could use it for

You don’t need to be a musician to want music that fits your project.

A small business owner putting together a promo video could describe the mood instead of picking a generic track: something fast for a fitness studio, or something quieter for a restaurant.

A parent making a birthday slideshow could try a soft instrumental. A student working on a short film might want music that builds tension toward a big moment. Someone who writes poetry could hear what their words sound like set to a melody and sung by a synthesized voice.

Projects can be as simple as background music or a lot more involved. With compatible software, an AI agent can help write melodies, suggest arrangements, change instruments and try out different sounds. You could ask for a slower version of a song or piano added to an arrangement you already have, and keep refining it in conversation instead of starting over each time.

ChatGPT needs the right software

ChatGPT doesn’t turn into a music studio just because someone asks it for a song. It has to be connected to software built for writing and editing music.

Some developers are already testing MCP connections for professional music apps. ACE Studio, for example, supports connections with compatible AI agents, so some production tasks can be handled through typed instructions.

This works differently from the usual AI music generators, which take a short text description and return a finished recording. With connected software, you can work on separate parts of a song, ask for specific changes and shape the result step by step.

Not every app has every feature, and some connections take technical setup. Still, instead of learning every button and menu, people can increasingly just say what they want done.

Music for people who aren’t musicians

Cities with lots of small businesses and working artists are full of people who’d use this.

Plenty of people make digital content without thinking of themselves as musicians. Local restaurants post promo videos. Real estate agents film property tours. Families record vacations and celebrations. Students make films, and artists look for new ways to show their work.

Many of these projects would be better with music, but hiring someone to write an original soundtrack, or learning complicated production software, often isn’t realistic.

AI-assisted music tools could be another option. Instead of relying only on existing recordings, people could try sounds that suit their project better.

That doesn’t mean every business is about to write its own advertising jingle. But more people will get to try ideas that used to seem too complicated or expensive.