How AI and TikTok Are Breaking the Music Industry - Rick Beato

Rick emphasizes that the old model of selling millions of CDs is gone, and even million-stream songs yield relatively small payouts. He points to his friend Tosen Abasi (Animals as Leaders) who has a guitar company, plugin sales, and several income pillars, and Polifia's Tim Henson with a signature Yamaha strap and Neural DSP model. He also mentions VIP meet-and-greets before shows and group lessons as profitable add-ons. The key is to treat your career like a small business, constantly testing and adding revenue streams. He frames this not as optional but mandatory: 'You have to figure out what your recipe is for making a living.' Even a 63-year-old like himself learned to navigate this new landscape, so younger artists have no excuse.

Rick explains that he trained two separate voice profiles — one for his shotgun condenser mic (4 feet away) and one for his SM7B close mic — because the AI initially gave him a close-mic voice for a video that used the distant mic, ruining the blend. He has used it to secretly replace lines in B-roll and reports that no one noticed. He also uses it non-publicly: this morning he had some written ideas he wanted to review, so he asked his assistant to feed them into 11 Labs so he could listen to them read aloud. He notes the voice verification step is smart: after training, you must read a paragraph in real-time to prove it's your voice, preventing theft.

Rick suggested that mastering is an area where AI can genuinely add value because it's more technical and less creative than mixing. He went further: if you could prompt an AI to mix your song like Chris Lord-Alge or Brendan O'Brien, you could get a useful starting point in seconds. He acknowledges mixing has 'taste' but sees this as a tool for exploration rather than a threat.

Rick explains that in today's pop and country music, tracks are often built first by producers ("track guys") who program drums, guitars, and other instruments. They bring near-complete backing tracks to writing sessions where multiple songwriters from different publishing companies contribute lines, melodies, and lyrics. The artist may come in at the end to sing, sometimes with zero involvement in the writing. He contrasts this with the 1980s when even artists who worked with co-writers had a more hands-on role. The shift towards producer-driven music means the producer's sonic signature often matters more than the artist's creative voice, leading to a homogenized sound across radio. He gives the example of Kelly Clarkson's 'Since U Been Gone', which was entirely written and produced by Max Martin and Dr. Luke before Clarkson sang it.

Rick observes that the old system of record labels building a career through radio is dead. He describes the chain: a song has to have a viral 'moment' (like the memes around Bring Me the Horizon's 'Can You Feel My Heart' or Fleetwood Mac's 'Dreams' with the Ocean Spray skateboarder) to break through. This two-way dynamic means artists now intentionally design songs with short, high-impact sections suitable for TikTok clips. He cites the song 'Headlock' by Imogen Heap (released 2005), which resurfaced in 2024 after multiple anime edits used different parts of the song, causing a massive spike. He's actively studying apps like ChartMetrics to reverse-engineer what makes a song TikTok-viral, concluding that the formula is still partly unpredictable but heavily data-driven.

Rick details a current case: a band called Velvet Sundown appeared on Spotify with a verified checkmark, AI-looking images, and a full album. Despite zero history or public identity, it skyrocketed to over 600,000 followers. A second album is already announced. The music doesn't sound good to him, but the quiet integration suggests Spotify is testing audience tolerance. He fears the scenario where labels and platforms bypass artists entirely: if Spotify owns both distribution and AI artist creation, it can keep all revenue. The legal gray area is huge — if someone prompts an AI song, then re-records it with human musicians, who holds any copyright? He predicts eventual market segmentation: some listeners won't care if music is AI, much like graphic design was disrupted by Midjourney.

Rick notes that many country producers are former rock/metal engineers. Joey Moi, who engineered Nickelback, moved to Nashville and became one of its biggest hitmakers. Bands like Jelly Roll's crew were originally in death metal acts. This influx brought rock energy and production values to country, resulting in a sound that feels familiar to rock listeners. He contrasts this with the modern rock scene, which suffers from sonic sameness due to digital amp modelers and sample replacement, making it less emotionally dynamic. Country's live scene and authentic storytelling (even when factory-produced) filled the void. He points to Morgan Wallen's songs as examples of music that sounds more like pop than classic country.

Rick argues that because most guitarists now use digital modeling units, they are painting from the same sonic palette. This erases the happy accidents that came from physical amps and mics — like a mic falling to the floor and creating a new tone. Combined with drum samples and autotune, the result is a smooth but lifeless sound. He contrasts this with records mixed by legendary engineers like Andy Wallace and Brendan O'Brien in the '90s and early 2000s, which still sound massive and punchy because of the human touch and room dynamics. The loss of dedicated professional producers and mixers in rock music (due to shrinking budgets) exacerbates the problem, leaving bands to self-produce using YouTube tutorials.