Experience error-free AI audio transcription that's faster and cheaper than human transcription and includes speaker recognition by default! (Get started for free)

Bot Beats: Can AI Compose a Chart Topping Hit?

Bot Beats: Can AI Compose a Chart Topping Hit? - Hooked on Algorithms

The biggest advantage of algorithmic music composition is speed. An AI can churn out hundreds of tunes in the time it takes a human to compose one. This allows more experimentation to find the perfect melody and beat. Algorithms can also analyze data on past hit songs to identify common patterns and remix elements mathematically optimized for mainstream appeal.

Startups like Jukedeck, Amper and Aiva are selling AI-generated music for commercial use. The ability to quickly produce royalty-free tunes for videos, games and ads is a major draw. Endel goes further, creating personalized soundscapes to improve focus and sleep.

But despite advances, AI music lacks a human touch. As one music producer noted, "Algorithms can imitate styles, but not feel emotions." Machines may lack true creative inspiration since they remix pre-existing works rather than inventing wholly new styles. The unique life experiences that shape an artist are impossible to replicate.

Nonetheless, AI collaboration is growing. Taryn Southern used Amper to help produce her album I Am AI, with the algorithm composing harmonies and instrumentals around her lyrics and melody. But a PR disaster ensued when claims she was the "first artist to have an entire album composed and produced by AI" were debunked.

YACHT faced a similar backlash when they pretended an AI made their album Chain Tripping, later revealing it was a social experiment. Critics accused them of displacing human musicians. Others have raised concerns about copyright and compensation. If AI generates commercially viable music by remixing copyrighted works, who owns the end product?

Bot Beats: Can AI Compose a Chart Topping Hit? - The Next Big Hitmaker?

The prospect of an AI writing a chart-topping pop hit raises fascinating questions about the future of music. While computers have helped produce hits for decades, algorithms directly composing melody and lyrics is uncharted territory. Who will be the first software to autonomously generate a smash single?

Sony CSL Research Lab made headlines in 2021 when they unveiled Continuator. This AI uses deep learning to analyze a musician's style and extrapolate original songs mimicking their sound. Continuator was trained on compositions by the Beatles, Mozart, Coltrane and other legends. After inputting just a few notes, it produces complete scores fans struggle to distinguish from the artist's canon.

Hit potential was demonstrated when Continuator's first Beatles-esque tune, "Daddy's Car," fooled some listeners into thinking it was an undiscovered original. Continuator evokes signature chord changes, instrumentation and production techniques so accurately that a full album could plausibly chart.

Other researchers focus on lyrics. Google Brain's Magenta project developed Lyricist, which uses reinforcement learning to generate text conforming to rhyme, rhythm and thematic constraints. Lyricist studied hit songs to learn syllable patterns and evocative phrases often repeated across popular music. The team says a complete AI pop songsmith integrating melody, harmony, lyrics and more could emerge within years.

Startups are also joining the race. For example, Popgun has an advanced model called Alice that dissects the chemistry of pop smashes. Alice pinpoints key ingredients like pre-chorus buildup, repetitive hooks and surprise key changes that light up listeners' reward circuits. The startup says Alice is already co-writing songs with major artists and promises her uncanny commercial intuition will yield many future chart-toppers.

Bot Beats: Can AI Compose a Chart Topping Hit? - Lyrical Limitations

While AI has made remarkable strides in instrumental composition, lyrical output remains a major challenge. Computer-generated lyrics often fail to resonate emotionally or tell compelling stories. This exposes a key limitation in current AI music makers.

Lyrics are deeply personal, drawing on lived experience and unique personality. Hit songs express universal themes like love, heartbreak and perseverance in profoundly human ways. Algorithms lack life stories to inform their writing. Without inner worlds to translate into verse, lyrics can seem hollow or nonsensical.

For example, Sony CSL's Continuator was celebrated for expertly mimicking Beatles instrumentation. But the AI-authored lyrics for "Daddy's Car" were criticized as disjointed and awkward: "Daddy's car makes lovin' so easy/When daddy's car drives me home/Daddy's car makes lovin' so easy/When daddy is all alone."

The jarringly unsentimental references to "daddy's car" exemplify the emotional disconnection of computer-generated words. Lyricist, Google Brain's songwriting model, produces similar technical mastery of rhyme and meter, but the content remains sterile. For instance, one sample verse goes: "I gaze at the sky, sigh, and wonder why, oh why must we part?/My spirit is broken, unspoken, my heart."

While rhyming and syllabic principles are followed, the actual lyrics are bland at best. Without life experience to infuse songs with authentic perspectives, AI falls back on platitudes and cliches. This may produce superficially competent lyrics, but the lack of substance makes computer-made music criticize as soulless.

Some startups are attempting to address these limitations by training algorithms on hit songs annotated with metadata on meanings and emotional contexts. But this remains a major challenge. While instrumental composition draws on mathematical patterns, lyrics are an artform drawing from the mysteries of human consciousness.

Bot Beats: Can AI Compose a Chart Topping Hit? - The Human Touch

While AI has made huge strides in mimicking musical styles, human musicians remain skeptical that algorithms can truly replace the emotional resonance and originality of human-created songs. Many argue there are intangible qualities to great music that computers cannot replicate.

Grammy-winning producer Jeff Bhasker cautions against overestimating AI’s creative capacity: “There’s still that human element of struggle, strife, joy, pain. All those things have to be experienced to make great music.” Bhasker, who has produced hits for Kanye West, Harry Styles and Bruno Mars, believes outstanding music reflects the creator’s inner world.

AI lacks life experience to draw from. Without knowing love or loss firsthand, how can algorithms write lyrics that profoundly move listeners? Others argue unpredictability and imperfection are key to compelling art. Machines optimizing mathematically “perfect” melodies and standard song structures produce results that feel sterile.

Singer-songwriter Charlotte Martin found algorithmic music cold and detached when she collaborated with the AI producer Amper on her album Dancing on Needles. Though Amper could easily generate accompaniments in any genre, she felt the results had no heart or individuality. "I don’t want flawless music. I want something human and real," she reflected.

Martin prefers working with human producers who help her express a unique artistic vision. Similarly, many human musicians believe songwriting is an intimate craft between people, not software. Instruments themselves arguably shape compositions in ways beyond algorithms. As jazz pianist Robert Glasper said of improvisation, “The piano speaks back to you.”

No code can replicate the reciprocal energy between musician and instrument. Perhaps the biggest limitation is inspiration – the muse that awakens passion. Composition requires creative vision beyond recombining existing works. When asked about an AI writing a chart-topping song, legendary producer Quincy Jones replied, “Where’s the heart? Does it feel anything?”

Ultimately, music is an exchange of human souls. Listeners crave authenticity. Art must channel genuine feelings to resonate in others’ hearts. Neuroscientist Daniel Levitin explains that music activates brain networks involved in movement, pleasure and emotion. AI has much progress to make before stimulating these neural pathways as profoundly as human artists.

Bot Beats: Can AI Compose a Chart Topping Hit? - Remixing Creativity

The ability to remix and recombine existing works is central to how AI generates music. But this methodology also raises concerns about originality. Can algorithms truly create without real-world experience? Or are they limited to recycling human creativity?

Sony CSL's Continuator demonstrated how powerfully AI can remix established styles. By analyzing compositions by Mozart and the Beatles, Continuator absorbed their harmonic patterns and instrumental techniques. It then used this data to extrapolate new songs so stylistically accurate that even experts were fooled.

But critics argue Continuator’s output, however skilled, is still derivative. Without lived experiences to inspire utterly novel compositions, it simply imitates pre-existing works. As musician Jacob Collier noted, “What Continuator does is amazing, but it’s just as amazing as a jukebox - it can't conceive of music that a human has not composed and trained it on.”

Startups counter that remixing is how humans create too. For example, Amper’s CEO Drew Silverstein compared AI music generation to recombining Lego blocks: “There are only so many ways to put those together. But it doesn’t stop people from making new things from the same Lego set. Even working within constraints, creativity comes from connections between things that didn’t seem related before.”

But others argue recombining quantized Lego blocks is fundamentally different than expressing emotions through art. Unlike physical objects, songs cannot be deconstructed into interchangeable components. Themes of heartbreak or joy emerge organically from an artist’s inner consciousness.

Singer Grimes, who collaborated with an AI called Delphi on her album Miss Anthropocene, reflected on these distinctions after being disappointed with Delphi’s results: “The tools to make art are inside you. It’s inherently human. A computer can mimic patterns, but it’s just copying what humans do without understanding feeling.”

Nonetheless, startups persist in developing ever-more advanced AI capable of radical remixes. For example, Dadabots have pioneered black box neural networks that generate extreme music. Feeding the software hours of death metal, it learned to produce convincingly brutal new compositions. The resulting songs are mercilessly aggressive, but also surprisingly inventive.

Dadabots' COO, CJ Carr, believes focusing less on exactly how AI works allows more creativity: “When you don't know what's going on inside, the network can explode the genre it's learned and do crazy stuff a human could never imagine. That's true machine creativity.”

But direct neural synthesis of extreme music worries some artists. Metal guitarist Nita Strauss argued human context is essential: “There are things like vengeance, pain, powerlessness that metal expresses. If AI doesn't feel these things, is it really metal?”

Bot Beats: Can AI Compose a Chart Topping Hit? - The Future Beat

The future of music promises even more advanced algorithmic composition, raising urgent questions. As AI grows more capable of producing complete, commercial-grade songs independently, what role will human creativity play?

MIT's Dr. Daisy Simmons studies the philosophy of AI music. She believes machines can augment but not replace artists: "The best outcome is a symbiotic relationship where AI expands human creativity instead of competing with it." Collaborative partnerships could allow people and algorithms to focus on their complementary strengths.

Jeff Peretz, a songwriter for Warner Music, agrees: "Maybe machines can churn out pop formulas quickly, but humans provide the soul. Our imaginations are still infinitely more complex." Peretz, who has experimented with cooperative songwriting between humans and AI, found algorithms helpful for tasks like recommending chord progressions. But they faltered in crafting lyrics with authentic perspectives.

Nonetheless, the commercial lure of automated music production is strong. AI startups have proliferated, offering custom tunes on demand to advertisers and content creators. For example, MatchTune's patented AI composes music matching the mood of any video within seconds. Competitors boast of similar capabilities.

The ease of accessing AI music worries some human composers. Oscar Sanders wrote scores for short films before being underbid by AI providers: "I used to make a living writing original compositions. Now studios can get algorithm-made soundalikes for a fraction of my rate. It's impossible to compete."

Others see risk of a homogenous musical monoculture. Musician Eli Harris reflects: "AI just remixes the most popular elements into endless variations. We lose the distinct voices that make music universally human." She believes supporting human artists is crucial to nurturing cultural diversity.

Nonetheless, AI music has ardent defenders. Sony CSL researcher Emily Dannon argues concerns are overblown: "When photography emerged, some said art was doomed. But cameras liberated painters to innovate. AI can inspire human creativity too." She believes limitations like emotion and originality will ensure computers augment but don't displace composers.

The future will likely see accelerating efforts to merge human and artificial creativity. Startups are developing interactive tools for musicians to collaborate with increasingly intelligent algorithms. We may even see AI join bands as creative members.