It was only a matter of time. With tools like Google NotebookLM’s podcast feature, anyone can instantly generate an entire podcast on virtually any topic—no host, no production, no discussion required. Just feed the system some text, and it outputs a smooth, synthetic voice reading algorithmically distilled summaries. The result sounds like a podcast. But it isn’t one—not in any traditional sense.
Take @BooksPodcastco, a YouTube channel that presents itself as a series of literary podcasts focused on recently published books. At first glance, it looks legitimate: pick a book, hear a short episode. But listen closely and the illusion breaks. Every episode is voiced by the exact same synthetic narrators—identical cadence, tone, and rhythm. That’s because every episode is auto-generated using Google NotebookLM’s Audio Overview feature which takes documents in a given Notebook and outputs an episode of the “Deep Dive” podcast where two AI personalities discuss the materials. And nowhere on the channel is this disclosed—a direct violation of YouTube’s terms of service, which require creators to clearly identify AI-generated content.
Yet the deeper issue isn’t just regulatory. @BooksPodcastco is a small but telling example of the broader shift we’re witnessing: synthetic media not merely mimicking the real, but replacing it.
You can find the YouTube channel here (while it’s still up): https://www.youtube.com/@BooksPodcastco
