How AI Will Impact the Music Industry
AI is a major existential risk to the music recording industry
By artificial intelligence (AI) music, we mean the production of new songs that, primarily, will be different from current songs in four fundamental ways:
- CREATIVITY — an exponential expansion of artists’ or anyone’s creative abilities, far beyond their honed styles.
- SPEED — write and produce songs in a fraction of the time it takes today.
- COST — produce songs at a fraction of today’s comparable product cost.
- SCALE — the volume of music content AI can produce over a given time period is exponentially larger than that of non-AI music.
“AI will be able to do everything better than us humans.” — Elon Musk
This new technology will be in the hands of artists, fans, distributors and labels at different stages of adoption and levels of sophistication. Subsequently, major music record labels (Universal, Warner, and Sony) and music streaming service providers (Spotify, Apple, Amazon, and Google) — across the industry’s value chain — will race to build AI music production departments.
At first, artist and fans, independently and jointly, will embrace AI assisted and autonomous music production. Then, shareholders of the major labels and streamers will be asking management — why aren’t you developing your AI faster? because your competitors are.
Today, we are at the dawn of this disruption. AI music production is now capable of writing composition in MIDI phrases, and assembling them in song structures. For example, at our startup, Humtap, we have had such capabilities for some time now. While we are pushing forward the music output quality, a few others are racing to achieve variations of products.
Where Will AI Music Production Go from Here?
I get asked this question on a daily basis by members of the music industry, artists, and fans alike. Artists will start to work with AI music-making technology to produce new material faster, cheaper and at higher volume — a few already started.
Music lovers and fans will endorse this tech, well before anyone else. They will create their unique songs and music content using apps like Humtap — available today for free on Apple’s App Store — and share and distribute them on YouTube and social media, including as copyright free soundtracks for the videos they share online.
The music industry will look radically different sooner than anticipated. Remember how the TV broadcast networks felt about YouTube just a few years ago? Today, teenagers and millennials watch YouTube more hours each week than all of broadcast TV — something the leadership at NBC, CBS or ABC would not have fathomed a few years ago. The ad revenue that goes to YouTube today has already been lost from these TV networks, and it is too late to bring it back despite the networks’ scrambling efforts.
It is time for collaboration and partnership between major labels, artists, and leading innovative startups to create the future of music. This collaboration will foster a much better music experience, a faster path to revenue, and more inclusion of musical artists than if startups go at it alone.
From a financial standpoint, the record labels will create new revenue streams by distributing new songs produced at a fractional cost. Artists will expand their fan bases by reaching new markets. For example, an artist could create songs tailored to different genre lovers in no-time, in collaboration with other artists or just in conjunction with AI — e.g., Coldplay or Depeche Mode ft. AI.
Recording artists’ cataloged song stems can easily be infused into new AI songs for music fans to connect more deeply at a creative level with their musical idols. Artists will free-up some of their resources which AI will replace, allowing them to focus on higher value music creation and production that only humans can do. Music content produced in ways that are tailor mixed for different distribution channels, e.g., YouTube vs. a future Tesla car music streaming subscription service, or Spotify vs. a social media app.
Recent Market Development
- The AI music sector is starting to draw mainstream attention
- Spotify has entered the AI music sector
- The other primary music streaming services are exploring the space
- The three major record labels are taking this seriously by starting to form their AI strategies
- Valuations are rising for AI music startups
The Guardian recently published an extensive article about the bright future of AI music-making, listing and mentioning several players in this new sector, including Humtap, Google, Sony, Spotify, and a few other startups in the US and the UK.
Another article reads, “Abbey Road Studios … [has] gathered a group of startup companies who work on the development of artificial intelligence to see if there is an answer to the question: can machines make music? By the developments so far, the answer seems a definitive yes. … Startups like … Humtap … have already begun to make inroads.”
The current media coverage of AI music is significant in three ways. It indicates rising mainstream attention for adoption — not only in the US and UK but globally. Humtap and its competitors have garnered press from publications around the world including China’s largest online news site, and a mainstream German news outlet.
Many other articles about AI music imagine a brighter future for the music industry and fans with the rise of this innovative technology. Last but not least, expectations around valuations are rising.
Several publications raise controversial questions such as, ‘Will AI ultimately take over the music industry, replacing human hands and voices?’ and, ‘Will AI eventually replace human composers and musicians?’ We have received a broad range of feedback from major label executives and prominent artists alike.
Humtap’s Unfair Technical Competitive Advantages
From where we stand, today, Humtap is far ahead of the pack when it comes to the core AI music-making tech.
We created the hardest and most challenging parts of this AI music production puzzle, and we did it with the artist and human creator in mind. Humtap is the only technology that delivers on the artist or fan’s creative visions through their voice, humming and tapping, as the core of the music. We built a single end-to-end platform for both music composition and production.
We have also reached a point where we can run our deep learning models to understand a particular (established) artist’s music catalog so that it creates new AI models composing melodies, drums, bass tracks, and others, in that artist’s distinct style. This breakthrough coupled with traditional instrument sampling methods means we will have new content that is artist specific.
We often get asked, what are the critical differences between the various AI music efforts. In summary, these are Humtap’s competitive advantages today, which are hard problems we have been working on solving.
- User’s Voice Input — Humtap composes music based on the user’s creativity. The user hums a melody and taps a beat, and the platform creates songs from that input alone. To date, none of our competitors take the user into account. Others in this space generate random music decoupled from the user or ask abstract questions like choose a music key or mood — far from natural human creativity for the masses. While some of these may produce beautiful music, these are not relating to any input from me the user! Our tech enables real interactivity, with blazing speed, and with virtually zero lag or latency, putting the user in control of what they are doing and changing the music they are creating on the fly.
- Speed — Humtap’s music generation and processing is hundreds of times faster than its nearest competitor. For illustration, comparing 1/10th of a second vs. 10 seconds of processing means a 100 times speed advantage. The rest of the competitors utilize similar technologies to what Humtap developed about three years ago, which we have since moved away from and wholly upgraded to give us this advantage.
- Mobile Delivery — Humtap’s complete software including jAI and rendering components are entirely mobile. The Humtap app is self-contained and does not require any network or Wi-Fi signal. By comparison, our competitors base their products on server software. This advantage is no small achievement in the sector, given the processing power required, storage space needed and the limitations of what is possible within the iOS environment — Android is our next challenge.
- Mobile AI — Our AI inference composition models are built and adapted to operate entirely within the constraints of mobile devices. The competition has yet to catch up in this area.
- Originality — Many of our competitors rely on pre-recorded music, audio or MIDI, and sample loops in their music output. Humtap creates an entire song on the fly — none of which is pre-recorded or copyrighted. This attribute enables us to build unlimited unique songs customized to the user.
- End-to-End Platform — Humtap is the only technology that offers an end-to-end platform for capturing a user’s creative idea, composing, arranging, rendering, mixing, and structuring a full song instantly on the fly across multiple music genres and styles. None of our competitors have anything like this today.
The Future Is Bright For AI Music
We are in the midst of a healthy dialog with executives across the music industry’s key players, from the major labels, artists, largest streaming service providers to the tech world’s giants, FAANG. We are highly inspired by what we have in the works. Our underlying principle is guided by the question: how will we make the experience spectacular for music lovers and artists alike?
Among our core objectives is helping the music industry and artists to grow and collaborate on new revenue streams and unleash new levels of creativity and user interactivity for consumers who are actively sharing content on social media. We are excited about this opportunity to partner with artists and across the industry to lead the next music revolution and creative movement.
We look for a future where music connects the world through unleashing the creativity within us.