ITOKA: Meet the man about to change music forever with AI: “Those not willing to adapt might suffer”
In 1976, Neil Spencer penned the headline, “Don’t look over your shoulder, but the Sex Pistols are coming.” It was clear then that something was afoot in the cultural zeitgeist as punk reared its gloriously ugly head. Now, the music world is on the precipice of another revolution—one of an entirely different magnitude and order, and we are sailing into it blindly.
ITOKA are a music company ahead of the curve on this front, and I sat down with their co-founder, Yihao Chen, to grill him on whether he is about to destroy the music industry or simply open up a new avenue of creativity. Far from some nefarious bigwig steaming into the next fad, what I find in Chen is a highly personable individual with the sort of lulling voice and gentle manner that instantly quells any interview caginess.
In keeping with his character, Chen is not wading into this venture with futurist abandon either, but rather a clear-eyed openness to the spiritual and philosophical implications. Nevertheless, alongside fellow co-founder Malcolm Yang, he is about to foist a revolution. Their company is dual purpose: “One is a web-based application that allows people to create music easier and faster with the help of our artificial intelligence. And the second is we are creating an ecosystem of music to allow people to better utilise the music content that they created using the AI.”
Essentially, it is a one-stop shop: you can visit ITOKA, use its AI functions to write a song in five minutes, then sort out the paperwork, and have the song on a platform all in one place. That, in short, is streamlining in a manner that is slicker than a penguin’s back. And if his vision even comes close to fruition, then the music industry will never be the same again.
To ease any thoughts that he might just be some rampant capitalist super-villain railroading the music industry towards a profitable apocalypse, he ensures me that the project is inspired by a musician pursuing a new creative frontier. “I was a classical flautist, and I really love music,” he says. “And before I went to college, I was amazed by the pioneering music AI work being accomplished by Google in 2015.”
Chen saw their technological breakthrough as eliminating hurdles to creative potential. So, he got cracking. Now, if you want to create a song with ITOKA, Chen says: “It’s very easy. At this moment, you just simply log on to our web application and go to our ITOKA engine, which is our AI-powered music creation tool. Then you would simply be asked to input some options. For example, we would ask you what kind of genres are you looking for; then we will provide you with several demos for you to choose from.”
Once you’ve selected the demo, it is time to polish it up. “Our AI will make the demos into a full song,” he says, “and you will be navigated to an in-house music customisation interface to do some customisations and make the music to your tastes. Afterwards, if you’re satisfied with the result, we will help you render that to a full song in audiofile. And that it’s the format that we have right now.”
That is wild enough, but they want to develop this even further. Chen explains: “Ultimately, we will enable several different modalities of inputs. For example, text to music. In the future, we you will be able to directly input a line of script for example, ‘I want a song for a rainy day sitting in a coffee shop in a moody environment for three minutes’, and voila, we will give you the song that you are looking for with just that simple input.”
Chen continues: “Also, you can imagine that this service can be integrated with other applications like game design engines, or content creation applications like Canva where we can get your pre-made posters or short videos and get you the best fitting music given that situation.” In other words, if you are a YouTuber or podcaster and you want some background music that will pair well with your true crime episode on sinister ministers, then rather than scour the web for something that sounds fitting and jump through costly copyright hoops, or even the rigmarole of recording your own tune, you can just instruct AI to do it for you. Of course, I put the implications of that to Chen too.
But in order to leap down the spiritual wormhole, it is important to first establish the parameters of the reverberations of AI in music. After all, it’s one thing to create a jingle for a TikToker, but it is a monumental leap to usurp the charts. “The possibilities are enormous,” Chen asserts. “Everyone who needs music or needs customised music can be benefited by our application. We can do almost everything you want with the music.” That includes the next generation of would-be superstars.
That notion of unlimited possibilities and being at the start of that journey is quite a terrifying proposition. In the last few months, the likes of Nick Cave and Noel Gallagher have all implied that AI heralds the end of music. But it is worth remembering that music is an art form with a very interesting history when it comes to technological interaction. As I put it to Chen, “When synths became popular, there was a musicians union strike against them in 1982. Do you see the creative future of AI as making the current naysayers seem as doom-laden as that strike against synthesisers now looks, and we’ll just incorporate AI into the musical process in the same way we have drum machines, sampling and so on?”
“This is a very interesting history in the music’s technological development,” Chen nods. “It’s very reasonable for people to be intimidated by the power of new technologies. And there might be some arguments or adversaries towards the adoption of the new technologies. But you already know that The Beatles successfully adopted synthesiser at the very, very beginning, using Yamaha synthesisers in a lot of their famous songs, and it was a hit! Now everybody loves it and admires the way that they adopted new technologies as a sound source in the music.”
This leads Chen to the most pertinent point of the current discussion: “The goal of this AI technology is not to replace human creativity, but to expand the territory or the boundaries of human creativity. In music, it looks at providing efficiencies.”
ITOKA also wants to “democratise this whole process. So that everybody in the world who is excited about music creation can have the same access as professional musicians to express themselves,” he said. “So, in a nutshell, I’m very optimistic about the final outcome of AI technology being adopted in the music industry. There can be arguments, there can be intimidations at the very beginning, but ultimately, people will gradually be accepting of the presence of AI technology in music production, and they will love it. They will enjoy the benefits that have been introduced by this cutting-edge technology.”
However, even progress can come with its pitfalls if it isn’t planned out properly. While levelling the playing field of music – especially in an age when it is increasingly being commandeered by the middle classes and access to necessary platforms is limited for working class artists – sounds like a great thing, there is a risk that the process is democratised to such an extent that the industry is suddenly flooded. In other words, if everyone can now make a little bit of money from the music industry, will it then make it harder for anyone to make a living?
“This is a very reasonable concern,” Chen agrees. “I believe that actually, it helps a lot with the development of the music industry, and changes how people monetise music. So, if you take a closer look at the music that were listening to on a daily basis, you will already see that there is a lot of isomorphic music content out there. Everybody is using the Canon D.” It is his rationale that the cream will still rise from the crop, but AI might help others on their journey towards it by providing quick income channels where currently there are hurdles.
This libertarianism, he claims, will also help to shake up the creative side of music. “Everyone is using the same chords, similar melodies, similar rhythm, same thing with synthesisers all the time, right? A lot of people are just getting lazy about being creative. I believe that one of the benefits of the power of AI is that we are pushing people forward when they are thinking about being creative in music, right? We don’t want to make people lazy. If you are making music similar to the quality of our AI technologies, you are all out of luck, right?”
He continues: “You have to make sure that you are putting a lot of effort and a lot of thoughts into making music so that you are standing out from the content created by AI.” While that might sound blunt, what Chen is essentially saying is that songs that are already partly regurgitated and half-cooked – the sort of shit we are forced to run away from daily – will be needled out by AI. Because, in essence, the rehashing of soulless popular music presently is no different to a slowed-down incarnation of AI anyway.
When the technology really takes hold, AI will be able to produce these basic songs in four minutes, forcing a new invigoration of differentiation upon the music industry. If you can’t beat the machine, then c’est la vie. But that doesn’t mean the machine will take over. We still love a hell of a lot of mono four-track tape recordings that wouldn’t pass muster from a technical standpoint now, but they work their way towards cherished additions to our daily lives by virtue of their originality and soul if all the technology between now and then hasn’t rendered them obsolete, then it is unlikely even something as novel as AI ever will either.
However, from this comfortable perch, Chen then sideswiped me with a statement that delineates the true revolutionary capacity of AI. So far, our chat had downplayed it as just an upscaling of a synthesiser with added benefits. Then he seamlessly announced: “The old ways might die”. Repeating for good measure: “The old ways of making music in the music industry might die… But new ways of monetisation will come forward. In my experience in the music industry, I don’t think the old way of monetisation is good. For over a century, the big labels, the royalties, licencing, etc, etc. AI can bring changes to this kind of inefficient way of monetisation and make people monetise the music content faster or in a more innovative way.”
On a practical note, Chen explains that if you were to create a song using ITOKA you would then hold the copyright yourself. “The short answer is yes; you can acquire the full rights or licences for monetisation for content, and you can even own the content with the help of blockchain technologies. We want to simplify the whole procedure for you. You can just go ahead with our infrastructure, and with a single click, you will be able to get the power or the rights that you want to have for your music content. Then everybody will have the same accessibility to the content and the monetisation.”
That might be all well and good for those creating music, but what about the existing musician’s work that AI is using to help generate new content? For instance, if you instruct it to create a drum beat akin to the work of Matt Helders or the piano work that inspires your café sounds request is lifted from Norah Jones, then what are the implications for them? Well, this is where AI gets very tricky indeed. “We are still looking for the optimal solution with that,” Chen concedes. This is perhaps why the world is so trepidatious about AI in general. Thanks to the rampant views of prominent libertarian advocates like Donald Trump and Elon Musk, we are used to powering on with progressive technology without much legislation. Now, we are on the brink of an AI revolution, and we’re moving ahead with it despite huge holes yet to be addressed.
Likewise, this has huge implications for the current big label system. And while it is widely agreed, as Chen says, that the current system is flawed and strips the artists of power, toppling it overnight could be even more catastrophic. Chen has a vested interest in that topple, but even he explains: “I believe that also in a very short time of period, people cannot abandon the existing music industry paradigm. It has to be a long-term transition, where we migrate from the existing one to a more optimal one, but where the optimal point is, we still have to investigate.”
Even though these legal issues might be huge practical hurdles, there are still those who uphold that the main issue is a spiritual one. If music is about human transcendence, then how can a robot really be part of the process and not result in us losing something along the way? “I’m a musician,” Chen explains. “And I saw the power of AI in terms of boosting people’s creativity and saving a lot of time on tedious music production tasks. I totally understand that there are some concerns about diminishing human creativity. But just like what happened with synthesisers or steam machines in ancient times, it’s just that it’s a tool, right?”
He continues: ”People can use this tool to do something more magnificent, something that can move our human societies and our cultural recognition further. They can bring more creativity into this space. Those people who are not willing to adapt and not willing to use this technology might suffer from the gap that we create between the people who are using data and who are not using it. So, personally, I believe this technology will help people in the long run instead of jeopardising human creativity.”
Ultimately, his sentiment is one that was also echoed recently by Billy Corgan. Speaking on the Zach Sang Show, Corgan, who has been an active musician since 1985, explained: “Once a young artist figures out that they can use AI to gain the system and write them a better song, they are not going to spend 10,000 hours in a basement like I did. They’re just not.” When asked about whether this will still result in “art”, the musician mused: “Ultimately, art is about discernment. Somebody was telling me the other day about how a famous rap artist would work. They would bring in all these different people and they would sort of pick the beat that they were most attracted to.”
He likened this process to a mere primitive form of a musician using AI as a muse for their craft. “Now, let’s change that to AI: ‘Hey AI, give me 50 beats’, listen and meh not really feeling it. ‘Hey AI give me 50 beats from the 50 most famous rap songs of all time’. Okay, I like number 37, that inspires me,” he hypothesised. Concluding: “Are they ripping it off? Not really because I did the same thing, I just did it analogue. I listened to 10,000 songs and I was like, ‘I like that beat’, so what’s the difference?” So, it is more a case of rapidly streamlining the current music process rather than usurping it.
But that in itself might be problematic for some. Isn’t the struggle of sitting in a basement grinding out 10,000 riffs and waiting for the inspiration part of what imbues a song with a detectable sense of sincerity? This is evidenced when I ask Chen if AI can really create the next Bob Dylan, and he says, “Yes, but we don’t want the next Bob Dylan. We want the next say, Tom Taylor. You can be the superstar.”
“There is a comical irony to that, though,” I argue back. I know my own musical capacity is limited, and I’m not fit to be a star. Granted, on a wider scale, you could argue that my working class standing has presented barriers as it has done for millions of aspiring musicians, and the hard work needed to overcome them without a leg-up has become unsustainable. But overcoming them in five minutes is perhaps too far the other way. While the cream might still rise to the top, the worry is that the path there will be crowded with a few too many Tom Taylors to really let the great musicians break through.
As Chen concludes: “A lot of music will be coming out and people will be flooded with it. People might be concerned that we will be losing the standard of defining good quality of music. I personally believe that this will not be the case. Because just look at what we are listening to right now. We have different kinds of functions of songs, right? There’s music that we appreciate as high quality music, like Mozart, Bach and timeless classics. This hasn’t been eclipsed by pop music or other kinds of music.”
He adds: “So, what I am trying to point out is that we human beings are can still establish a new paradigm or standard for us to judge whether music is high quality or not, we will still have a standard for that. And in the meantime, we will be using music in more ways that we haven’t thought about to maximise our creative activity.”
