‘I realised machine learning could make my musical dreams come true’

22 Feb 2019372 Views

Share on FacebookTweet about this on TwitterShare on LinkedInShare on Google+Pin on PinterestShare on RedditEmail this to someone

Tech innovator and singer Dr Maya Ackerman. Image: Shawn Flynn

Share on FacebookTweet about this on TwitterShare on LinkedInShare on Google+Pin on PinterestShare on RedditEmail this to someone

Tech innovator and singer Dr Maya Ackerman sees AI as the perfect testing ground for music, where people’s creativity can really flourish.

While there are many facets of artificial intelligence (AI) that seem destined to take over our lives, it seems that few pursuits are as likely to be taken over by robots as music. With the meteoric rise of music streaming and its ability to track our music interests, likes and dislikes, music producers have as good a picture as ever of what to make that has a high-percentage chance of topping the charts.

However, away from the business end, some researchers and artists are finding ways to use machine learning to create a human-robot collaboration that few would discern is based on an algorithm. One such individual is Dr Maya Ackerman, a leading AI researcher based at the Department of Computer Engineering at Santa Clara University in the US.

With dozens of peer-reviewed publications to her name, her research is focused on the concepts of computational media and computational creativity, those being the ways we can use machine learning in creative spaces as well as the systems that drive that creativity.

Speaking with Siliconrepublic.com, Ackerman said her interest in the intersection between AI and music started early on in her life, describing the first sight of her computer following her instructions as “euphoric”. However, at the age of 27 and in the middle of doing her PhD in computer science, she said something changed in her.

“All this time I was sorely aware that something was missing; the artist in me was starving, so I decided to start taking vocal lessons. This changed my life. Despite being a fairly awful singer at first, within nine months I became a semi-professional opera singer.”

At the age of 35, things changed drastically again when she took the massive leap to juggle both an academic career as well as being an entrepreneur as co-founder of a start-up called WaveAI. Through the start-up’s app called Alysia, a singer who is having trouble writing music can seek help from the app’s AI.

‘We listen to music to hear what a human being is expressing’

“As a musician, you really need to have your own music. But even though I worked hard on it for three years – learning to be a producer, taking piano and improvisation lessons – I simply didn’t seem to have what it takes to create good original songs.

“Eventually, I realised that my machine-learning skills could make my musical dreams come true.”

The app has recently launched on the Apple App Store and has so far seemingly caught the attention of a number of media outlets and, most importantly, singers. In the future, the company plans on letting users also generate their own music videos based on the songs they and the AI have created.

So, is this the beginning of the end for the human musician? The concept has long been mused by futurists for when there exists technology that can make humans sing better, so why not just create AI that can do everything?

According to Ackerman, this is a conversation far older than what we know of as AI today, with examples such as the acoustic piano once being considered as taking too much musical creativity away from the pianist. Not only that, but automated music as a concept could be dated back to 1960 when a Russian computer scientist and composer published the first ever paper on algorithmic music.

“Music is social; it is how we connect with each other. It is a deeply fulfilling form of self-expression,” she said. “Computers can support us but, ultimately, music will move forward through us using technology, not through the technology leading the way for us. We listen to music to hear and feel what another human being is expressing.”

Black and white image of Maya Ackerman singing into a microphone wearing headphones.

Dr Maya Ackerman singing. Image: David Loker

‘Far too often, I’m the only woman in the room’

If you think this sounds like a lot for a working academic and a CEO to take on, you wouldn’t be the only one. While Ackerman admits that what she is doing “would certainly earn a high score on any difficulty scale”, she said combining both roles has come natural to her.

“It became clear to me that Alysia – which started off as a research project – needed to be released into the world,” she said.

“It had to make it into the hands of users so that they could benefit from this breakthrough technology, and be able to express their feelings and experiences through original songs. It wasn’t right to keep such powerful technology solely within the rather opaque walls of academia.”

Certainly, aside from combining two tough roles, Ackerman is bucking the trend when you consider the overwhelming numbers of men – and particularly white men – in the computer science and AI space. The number of women in her space, she admitted, is “alarmingly low” and added that had Facebook, Google or Microsoft been founded by a woman, “perhaps it would not have taken me 35 years to realise what I want to do when I grow up”.

“When I founded my company, I was shocked to discover that the numbers are even worse in tech leadership amongst founders and CEOs. Far too often, I’m the only woman in the room,” Ackerman said.

“I hope to see massive changes in my lifetime, and I hope to lead by example, showing that there is a place for women in science and technology, as C-level suite leaders, inventors and innovators shaping the future.”

Want stories like this and more direct to your inbox? Sign up for Tech Trends, Silicon Republic’s weekly digest of need-to-know tech news.

Colm Gorey is a journalist with Siliconrepublic.com

editorial@siliconrepublic.com