Five minutes remain before taking the stage. A private event client suddenly requests a bespoke song the band has never rehearsed. Musician and sound engineer Ramil Kapushev prompts ChatGPT to generate the chord progression, harmonic intervals, and lyrical cues, reviews the output in 30 seconds, and steps up to perform. In his working life, artificial intelligence is not a terrifying existential threat, but an indispensable digital co-pilot for high-pressure routine workflows.
At 35, Kapushev maintains an active footprint across Astana’s music industry: playing in prominent live collectives (Sana Sezim, PIONY cover band, Hola live band), working as a remote session recording artist, mixing audio tracks for commercial clients, and teaching advanced music theory.
Generative Audio Tools: Suno, Udio, and the End of Stock Music
The music sector has experienced an onslaught of generative audio platforms: Suno v3 and Udio generate broadcast-quality full-frequency songs complete with vocals, instrumentation, and arrangement in 20 seconds from a single text prompt.
«Generative AI has undeniably eliminated the bottom tier of commercial background audio,» explains Kapushev. «If a boutique cafe, real estate agency, or mobile game developer needs generic ambient acoustic guitar or upbeat corporate lo-fi, paying a freelance sound designer $200 makes zero economic sense when an AI model creates ten variations for $10 a month. That market segment is permanently gone.»
Why Live Performance Remains an Unconquerable Economic Moat
While studio audio generation has become cheap and ubiquitous, the economics of live performance in Kazakhstan have actually strengthened. In a culture deeply anchored in corporate galas, family weddings (toi), music festivals, and intimate club gigs, synthetic tracks cannot replace the visceral human energy of charismatic musicians performing live.
«People don’t pay money at a concert simply to hear frequencies emanate from speakers,» says Kapushev. «They pay for eye contact, stage improvisation, athletic musicianship, shared emotional vulnerability, and community. An AI model can write a mathematically flawless jazz solo, but it cannot sweat under stage lights or read the mood of an auditorium.»
Streaming Royalties vs. Real-World Revenue Streams
Kapushev dismantles the romantic myth that independent musicians survive on Spotify, Apple Music, or Yandex Music streaming royalties. In Kazakhstan, streaming payouts represent fractions of a tenge per listen; unless an artist generates hundreds of millions of global streams, digital royalties barely cover monthly studio maintenance.
The authentic commercial engine for Kazakhstani musicians rests on three tangible pillars: private event bookings and corporate tours, session recording fees for major media productions, and private masterclasses and pedagogical mentoring.
«AI isn’t destroying musicians’ earnings—extreme market competition and lack of business literacy are,» concludes Kapushev. «Artists who learn to leverage AI to automate track stems, master EQ curves, and draft chord sheets will produce higher-quality music faster. The artists in danger are those who refuse to adapt, expecting the market to reward mediocrity simply because it was manually composed.»
