What Will Be Is Not What Should Be

The ocean waves are constantly in motion, sometimes roaring and crashing, other times gurgling softly. Even in chaos, there’s an underlying stability to the cycle. People reshape the sand every day–building sandcastles, digging holes, leaving their marks behind. Just as reliably, the tide wipes the sandy slate clean. The only way to preserve a sandcastle is to build far enough away from the water. But when sea levels rise, what once was safe is no longer out of reach.

Technology and innovation advance in a similar way, relentlessly wiping away what was and replacing it with what is. Older ideas, tools, and systems–and the people whose livelihoods depend on them–are bulldozed by whatever replaces them. The rallying cry is “progress,” and destructive effects are minimized in favor of promised benefits. Those who raise concerns are dismissed as anti-progress–reduced to a one-dimensional obstacle rather than a participant in the conversation. This was most visible in the podcasting industry, where AI-generated podcasts began flooding the ecosystem. In 2025, podcasters were suddenly confronted with the reality that they had built too closely to a rapidly rising sea. AI was indifferent–poised to erase the creative ground they had built on, leaving them to adapt or be drowned.

The Perennial Problem of Discovery

My daughter is about to graduate from the Fashion Institute of Technology with an illustration degree. She’s a creator through and through. But like most creators, it is daunting to graduate and then try to stand out in a crowd of illustrators. In a world in which AI can happily “replicate” most styles, she is not only competing against her peers, she also has to compete against the untiring bots that never paid for school, never needed to work hard, and never stop producing.

Podcasters are increasingly encountering a similar situation. Massive conglomerates (Apple, Spotify, SiriusXM, YouTube, and others) have been consolidating the landscape over the past few years. Additionally, technologies wielding the power of AI are flooding the podcast market with content. According to Listen Notes, nearly 30% of the 14,221 podcasts created in November were identified as AI-generated–content designed not to be heard, but to exist. Inception AI claims they are churning out 3,000 AI-generated episodes a week and managing around 5,000 AI-generated shows altogether.

As Techdirt eloquently wrote, “Flooding the zone with just an endless parade of human simulacrum isn’t going to do great things for the Internet’s already hugely problematic signal to noise ratio, or the public’s ability to differentiate the wheat from the chaff.” AI-generated content doesn’t just compete with human creators; at scale, it overwhelms discovery itself, drowning out human-created content through sheer volume.

Curators Will Save Creators

The algorithm was meant to save us from this overwhelming flood of media. But instead of surfacing quality content, it regurgitated whatever was optimized for engagement and retention rather than originality. It was never built to evaluate quality or originality–only engagement. Content is being produced at effectively unlimited scale, tailored precisely to what the algorithm rewards. If a human never encounters other human-generated content, discovery never happens.

This is why human curation isn’t a nostalgic return to the past; it is critical to the future of discovery. When algorithms promote noise, trust becomes the necessary filter–one that cannot be manufactured at scale and must be earned over time. That trust already exists. According to Acast’s research with Nielsen, “Podcast hosts aid discovery more than search and [word-of-mouth] recommendation” with podcast hosts being “almost trusted as much as friends/family when it comes to podcast recommendations.”

The oceans are rising; what once was safe is now at risk. Human-led curation must take precedence to ensure human-generated content survives.