I was pro-AI until it started sounding a lot human — I'm worried about the future

Credit: Lucas Gouveia / Android Police By Irene Okpanachi Published 3 days ago Irene Okpanachi is a Features writer, covering mobile and PC guides that help you understand your devices. She has five years' experience in the Tech, E-commerce, and Food niches. Particularly, the Tech space allows her to geek out and share knowledge. Irene is a couch potato who loves gaming, singing, listening to music, and eating when she's not typing furiously on her laptop. Sign in to your Android Police account Add Us On Summary Generate a summary of this story follow Follow followed Followed Like Like Thread 2 Log in Here is a fact-based summary of the story contents: Try something different: Show me the facts Explain it like I’m 5 Give me a lighthearted recap A human being didn't sing my most replayed song this December. My sister's current favorite artist isn't real either. Those aren't sentences I ever imagined I'd type, but here we are. Artificial intelligence has slipped into my playlists and every angle of my creative space. Now, it refuses to leave. As someone who writes for a living, I understood the appeal of Generative AI when it first broke out. ChatGPT and other tools delivered the speed they promised. That excitement has since faded after scrolling through multiple faux-birthday photoshoots and lifeless captions on my mobile devices. I catch myself wishing we could rewind to when many things were still made by humans, for these very reasons. Related 4 simple ways to create playlists and find new music with AI AI takes the DJ booth A few months ago, my niece raved to me about a new app she'd discovered. I was working on my laptop and half-concentrating on her discovery. The app is Suno. It's on the Google Play Store and on the web. The free music generator creates full songs from text descriptions or from audio you upload. If you type lyrics, it treats them as a narrative spine to compose music around. The system arranges the lyrics itself if you leave the box empty or describe a theme instead. Songs appear in your library where you can replay, remix, or generate new variations. There's a discovery-style feed where users share AI-generated tracks, which is why machine-generated music is gaining traction in a format similar to social media. Within minutes, my niece had created multiple tracks that wouldn't sound out of place on streaming apps or mainstream radio. They were complete with verses and harmonies. She's only twelve, and yet had effortless access to software capable of mimicking years of vocal training and production experience. I remember brushing it off as I returned to my tasks. I never thought of it again. I thought I understood the implications because it couldn't get any worse. Apps of this nature are usually bound to attract scrutiny, then either get banned or regulated until they fade out. The music industry may look powerful from the outside, but it's built on an intricate balance of rights, credits, samples, and permissions that make it fragile. If it were that easy to make hit records, we would all be charting pop stars. I was wrong. Credit: Lucas Gouveia / Android Police | SkillUp / Master1305 / Shutterstock The Fate of Ophelia is an afrobeat cover of Taylor Swift's new song by the same name. It's just one of TikTok's trending sounds and is now also popular on YouTube. While there are still arguments about whether it's artificial or not, one video alone under the sound has crossed 200,000 views, and it has a dedicated dance. If you search deeper, you'll find other non-existent artists like Xania Monet, who has been featured on the Billboard radio airplay chart and amassed over 44 million official US streams. It's forced me to confront how unprepared we are for when machines supersede assistance and outperform us. In a traditional setting, making a song is a slow, layered process that stretches across months. Conceptualization begins long before production, after which the producer chooses an arrangement that suits the emotion of the song. Then instruments have to be selected carefully before recording, and a series of other processes. This timeline doesn't even include the emotional labor where there are doubts, multiple revisions, the vulnerability of sharing something personal, or the cultural context the artist brings with them. If AI can generate a song in minutes, it bypasses nearly all of this and compresses a deeply human process into mere output. The most unsettling part of AI artists and apps is their intentionality, or rather, the illusion of it. Music has always been powerful because it comes from somewhere. You can usually guess where a song comes from without being told. The culture embedded in it is frugal, in that it doesn't waste emotion, but draws from what already exists. Even when the song is simple, there is weight behind it because of the singer's infusion of history, language, struggle, joy, and different emotions. Now, you can just generate an Afro-influenced track without the system knowing anything personally about the culture it's borrowing from. You can have a spiritual song sung by a voice that doesn't believe in or even understand the deity it's praising. French philosopher Jean Baudrillard once wrote about a future like this. He said that representations would stop pointing back to reality and start replacing it. His book Simulacra and Simulation has never left my mind. It was one of the many warnings we should've listened to before the AI pandemic began. I'm coming to terms with the fact that AI is here to stay. But these days, I'm slowly detaching from intelligent features because they've crossed a line between assistance and dependence. When used carelessly, they're distractions that train you to outsource your thinking and judgment. I should know. I once fed my likeness into these systems for headshots and sought motivation from them. The more I stare at the responses now, the less of myself I see. I'm taking multiple steps to intentionally deprive these tools. Likewise, let some parts of your life remain inefficient on purpose. We've never needed machines to validate our actions and creations as much as we think.