
What happened
A WIRED reporter who enabled YouTube Music's beta 'Beyond the Beat' AI DJ heard 30-second spoken intros between songs, including one that mispronounced Charli XCX as 'Charli XXC' and another that described an Olivia Rodrigo remix with Chappell Roan before playing the original without her.
Why it matters
The reporter's complaint is not the synthetic voices, which sound convincingly human, but the commentary itself — lengthy platitudes that pulled the listener out of the music rather than adding to it, which is why the feature was switched off.
Summaries like this, every morning.
The reporter's experiment sits within a broader push by streaming services to build synthetic hosts into their apps. YouTube launched its experimental AI DJ in 2025, while Spotify first introduced a similar tool back in 2023 and doubled down on it in November by expanding the personalities available and adding more controls over the commentary. The reporter upgraded to Spotify's premium tier specifically to compare the two experiences, and found Spotify's synthetic commentary tighter and more tolerable than YouTube's — though neither was enjoyable enough to keep using.
The complaints are not about audio quality. Both companies' voices are described as decent and convincingly human; what turned the reporter off was the content. YouTube's DJ produced what the reporter calls a lengthy paragraph of platitudes with little substance, and in one case described an Olivia Rodrigo remix featuring Chappell Roan as unpacking the 'nuances of queer dating' — then played the original track without Roan. On Spotify, multiple voices leaned on filler words like 'you know' and 'like,' which the reporter found gave an uncanny-valley ick.
The piece is a first-person account rather than a broad survey, so it does not establish how other listeners respond. Spotify, for its part, claims in its launch blog that 80 percent of listeners come back to DJ week after week — a figure the reporter cites while noting they will not be among them.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
Otter.ai CEO Sam Liang says the AI notetaker now generates $100 million a year in revenue and has transcribed…

The developer of VocalRemover.one, which accepts MP3, WAV, M4A and FLAC uploads for browser playback before do…

Gigazine tested the Insta360 Mic Pro, finding its AI noise cancellation clearly made voices easier to hear out…

AssemblyAI achieved 70 percent cost savings on inference compute versus On-Demand pricing by running speech-to…

Earlybird-29M, a 29M-parameter English-only model whose 4-bit weights total 16.8MB, launched on Product Hunt o…

Mistral AI published Voxtral Mini 4B Realtime WebGPU on Hugging Face
