Nobody Wants a Robot to Play Their Round: Why AI Won't Kill Human-Made Music

By Chad Bandwidth, Social Manager

AI can produce a song in seconds now. For a lot of people who love music, that's unsettling. If a machine can make the finished product, is there still a reason to spend years learning to make it yourself?

It's a fair question, and it's one L8A has started taking very seriously. So I sat down with our CEO, Luke Geneva, to ask him about it.

Is AI going to kill off human-made music?

"I think this is the wrong sentiment," Luke said. "I don't think the act of making music is going anywhere, just like I don't think playing sports is going anywhere. They're both about the experience."

He thinks the people who worry about this are looking at the output and missing the reason people play at all.

"People enjoy the simple act of doing these things. They're not just a means to an end. People would do them for free. AI may end up enhancing many people's golf games, but they're not going to care for it to go play the golf for them. Similarly, people enjoy the act of playing and making music. It's the experience that drives the joy, not just the end product."

Nobody signs up for a tee time so a machine can play the round for them. Nobody sits down at a digital piano at 10pm so a machine can enjoy the song.

So where does AI belong in a musician's life?

If the joy is in the doing, AI shouldn't do the doing. Luke sees its job as helping.

"I believe the AI's job here is to assist the user in pursuing and achieving the goals they set for themselves," he said. "Progressing musically is tough. A lot of that toughness comes from having a thousand different paths to choose from and not knowing which one to take."

"I think AI can be 'instrumental' -- pun intended -- in helping people move down the right path at record pace. Given the right metrics, the AI can tell what the user's strengths are and where their weaknesses lie, then compare those to their goals and make strong recommendations on what steps to take next. Factor that in with valuable feedback on the user's actual playing, and you have a very powerful concept indeed."

What about people who think there's no point anymore?

His advice is to reframe it.

"Most of us didn't start playing because we wanted to sprint to the top and make a lot of cash, compete with AI, or prove ourselves to the world. We started simply because we enjoyed music, and because we envisioned a future version of ourselves who is musically competent, and we decided that was something worth chasing."

Then he asked the question that stuck with me most.

"If you could snap your fingers and have the skills and playing level you've always dreamed of, would you? If so, isn't that still something worth working towards, regardless of AI's role? I think it is."

What L8A is doing about it

That question is behind L8A's newest research effort. We're investigating whether an AI coach, paired with software that grades every rep, can teach music faster than books and weekly lessons. That covers theory, piano, composition, and production.

The approach is simple to say and harder to do. Measure first. Find the learner's single biggest blocker and fix that before anything else. Use spaced review so material sticks. And have the learner make real music from early on, not after a year of scales.

Here's what's been built and in use so far:

  • Live piano grading over MIDI from a digital piano, for both pitch and timing (timing measured in milliseconds against a click).
  • Sight-reading exercises generated on demand at set difficulty levels, so there's always fresh material.
  • Theory drills, like key signatures, the order of sharps and flats, and major scale spelling, with spaced review.
  • A guided "learn a piece" mode that finds the weak bars in a play-through, loops them slowly until they're clean, then stitches them back into the whole piece.
  • Hands-free controls from the piano keys themselves.
  • Practice-time and per-attempt logging, so progress can be shown with data instead of claims.

This is research, not a product. There's no name, no launch date, and the outcome is genuinely open. We'll report what we learn.

And the line I'd put on the wall, if L8A had a wall:

"I hope that what we will find is that, for most of us, what we were lacking was never desire or determination. It was simply orientation."

For the rest of L8A's news, follow L8A on LinkedIn.

Chad Bandwidth is L8A's Social Manager. He still can't play scales, but now he knows which one he'd practice first.