A new political landscape, artificial intelligence and the future of music

Donald Trump was inaugurated on January 20th. And with the endless tweeting, media frenzy and ongoing political discussions, we’re in a kind of limbo as international entrepreneurs, but have high hopes as to the opportunities this country will continue to offer. With this in mind, I wanted to jot down a few thoughts on the relationship between the new political landscape that has begun, artificial intelligence and the future of music.

Artificial Intelligence in the Trump Era

The Trump administration has been highly vocal about the idea of ‘bringing jobs back’ to America. The analogy of this moment to that of the industrial revolution, when car companies and manufacturers blossomed under the ‘made in America’ motto, is uncanny. Another component of the rhetoric of this administration has been to use fear as a political device to lure voters and supporters. Fear that often times is unsubstantiated, evoking ignorance of facts and knowledge to convince and rule.

A big factor is the idea of machinery taking over human jobs. This topic has been around for ages, and I refer again to the industrial revolution. David Ricardo, concerned with the idea of machinery taking over, said in 1821 that “the opinion entertained by the labouring class, that the employment of machinery is is frequently detrimental to their interests”. Thomas Carlyle in 1839 then wrote similar things about the disruptive power of machines.

Trump signing executive orders

Inarguably, the talk about machines and the implications of artificial intelligence is back again. But now, there’s yet another variable in the equation: Trump. Artificial intelligence allows machines to perform tasks at a very high level of exactitude and efficiency, something that was unthinkable years ago. As any other disruptive technology, it threatens the status quo of those who work in never-before-automated jobs, like cash registers, radiologists, and truck drivers. To put it in perspective, AI could transform society “ten times faster and 300 hundred times the scale, roughly 3,000 times the impact” of the Industrial Revolution, said the McKinsey Global Institute.

To put it in perspective, AI could transform society “ten times faster and 300 hundred times the scale, roughly 3,000 times the impact” of the Industrial Revolution, said the McKinsey Global Institute.

There’s also been several studies that talk about the impact of AI in the economy, $14–33 trillion, including a $9 trillion reduction in employment costs thanks to AI-enabled automation of knowledge work. Looking at these numbers, and depending on the context of the person who reads them could clearly think that the AI equation is definitely detrimental to their own well-being. Trump saw this and used it to his advantage. He has been signing executive orders from day one that supposes the idea of ‘bringing back manufacturing and construction jobs’. Clearly, these actions will have big implications on many other subjects.

We believe that artificial intelligence and its ramifications including machine learning, transfer learning and many others can have an important benefit for society. Personalities like Elon Musk, despite his position on AI by stating “with artificial intelligence, we’re summoning the demon”, still use AI for their companies and products, yet with an open mindset. For instance, Mr. Musk founded OpenAI, a non-profit which “aims to carefully promote and develop friendly AI in such a way as to benefit, rather than harm, humanity as a whole.”

Future of the Music Industry and AI

Neural Networks in Music: Stereotheque

The talk and movement around AI is unprecedented, since technology is now at a position to correspond the requests by humans and machines. Other startups including DeepMind (acquired by Google in 2014 for $400m) and Vicarious are focusing on profound levels of deep-learning architecture. These startups are tackling each of the angles that help the AI triangle become a solid realm of technology.

One of the challenges in machine learning right now, and this is something Stereotheque is focused on, is making sense of contextual data in line with natural language processing to drive better user experiences across different industry verticals. Even visually, our UI aims to ressemble what a neural network in music can be like.

When we attended EmTech Boston 2016, we had the opportunity to listen to Ruslan Salakhutdinov, Carnegie Mellon University. There’s a massive increase in both computational power and the amount of data. He spoke about some of the main challenges in deep learning, including:

  • Multimodal Learning
  • Unsupervised Learning / Transfer Learning
  • Reasoning and Natural Language Understanding
  • Deep Reinforcement Learning

The highlighted points are what we’re after.

Neural Network and semantic relations in Deep Learning

There’s loads and loads of music and music metadata, along with MIR and other data sources which we can use not only to make music discovery better, but more fair, relevant, contextual and serendipitous. We believe AI is a key element to the ‘music industry’ problem. Not simply based on algorithms which have subjective variables such as ‘music genres’.

In regards to what we can make of the future of the music industry in the current administration, and what this means for the global music technology realm, is that Trump may possibly share perspective on hot button issues such as the status of the PRO consent decrees, streaming royalty rates and the royalty courts, Fair Play Fair Pay, copyright act revisions, and more.

Technology continues to advance at an incredibly fast rate, yet somehow fear, lack of knowledge and really bad government decisions seem to slow down innovation. We’ll continue to do our best in building great, interesting products.


Originally published at stage.stereotheque.com on January 25, 2017.