You learned to play the guitar from Youtube? Cool, check out my orchestral piece…
Digital technologies already made music production more universally accessible. They are shifting some jobs while letting everyone get their hands dirty with generative music apps.
AI Music, a British tech startup using the slogan “Music that listens to you” is incubated by Abbey Road Red, Europe’s only music tech based incubation program from Abbey Road Studios and Universal Music. Studio’s aim is to join forces of the latest technologies of machine learning and music production. And their startup AI Music, as you can guess, is working on generating music the way you train it. It’s still in process.
Like many other AI products, they use machine learning algorithms. Instead of predefined rules, the algorithm gets trained on large data sets and comes up with mathematical representations of the patterns and correlations it finds. Then uses them to accomplish given tasks.
A previous start-up from the program was Uberchord; an intelligent guitar learning mobile app from Berlin that wants to change the way people learn music.
Their app uses chordal recognition, it corrects your mistakes when you play, gives you real time feedback like a teacher and combines this with adaptive learning -understands your skill level and makes a personalized practices for you- increasing the difficulty of the lessons as you improve. Of course, just like any other learning app it uses gamification techniques to motivate the user; goals to reach, daily reminders.
Cofounder of the app, Simon Barkow-Oesterreicher says that they are frustrated with the way musical instruments are taught. Good teachers take their students to a level that they can create music themselves rather than reproduce. So Uberchord is not about how to move your fingers at the right time but to understand how music works. Which he claims speeds up the learning process significantly.
Other notable labels such as Sony Music and Warner Music Group partly financed Techstars Music, another music related startup accelerator placed in Los Angeles. Amper Music and Popgun are it’s latest apps.
Amper, very similar to Jukedeck based in London, is an online artificial intelligence service that composes, performs or produces original music fitting for your content. You choose the mood, style and length, then customize and design the sound with tools that don’t require special training or musical expertise. An original soundtrack for your advertisement, app, podcast, game is made ready in a fraction of time. (I’ve used it, it’s simply very fast and fun to play with.)
Beyond doubt these services are extremely useful for people having their videos taken down for copyright reasons. After you choose the customizations, Jukedeck names your track (‘Berlin Abandon’ it was for my Drum&Bass track) and offers the royalty-free license to use it for $0.99 if you’re an individual or small business. It gets slightly expensive as $199 if you want to own the copyright of your click made track. Groov.AI and Humtap are also working on similar services.
Even though these create excitement in music circles, some experts working on composing with algorithms don’t seem impressed. Like Nick Collins who is the founding partner of Chord Punch, an independent recording company that has books on computer based music production and produces music with algorithms. “Computer based experiments in music production have been made since Push Button Bertha in 1956.” Collins says.
Music industry consultant Mark Mulligan suggests: “AI music is nowhere near being good enough to be a ‘hit’, but that’s not the point. It is creating 21st-century muzak.”
Like many revolutionary new technologies, artificial intelligence creates worries, like the less need for workers in the profession. For independent composers and producers who produce original music for corporate advertising companies, it is frightening to think that such an application can produce similar works in a shorter time and at a lower cost.
Music tech is looking at the issue as it is a natural phase in music history. These apps are not good enough to compete with human-made music and not currently aiming for that. They just open new paths for music production and for new musicians.
References
- Dickson, B.,(2017, December 23). AI Has Been Creating Music and the Results Are…Weird.
- Dredge, S. (2017, August 06). AI and music: will we be slaves to the algorithm?
- Uberchord: the personal guitar training app. (2017, March 27).
- Yapay zeka müziğin geleceğini değiştirecek mi? (2017, October 24).
