Core Event: Xiaoyuzhou’s First AI-Focused Event and the 2026 AI Trends Report
In late August, Xiaoyuzhou hosted its first AI-specialized offline event, ‘Signal Tower,’ and published the 2026 AI Trends Report. The platform has already implemented distribution restrictions on AI-generated bulk content behind the scenes, while explicitly framing its growth strategy around ‘strong and enduring’ rather than ‘small and beautiful.’
Key facts:
- 2025 metrics: 422% year-on-year growth in partner brands, 760% rise in commercial revenue, 356% increase in monetized podcasts;
- AI-related content: 187% growth in AI-focused creators and 239% growth in AI-related programs in 2025;
- User growth: ~50% in 2024, remained steady in 2025;
- Video podcasts: Under internal testing, positioned as an optional, on-demand feature;
- Voice updates: Launched two weeks ago, already used unexpectedly—e.g., hosts singing karaoke on the platform.
Commercial acceleration and ecosystem reconfiguration: Podcasts enter scale-up phase
Over the past year, podcasting has rapidly shifted from niche interest to mainstream attention. Bilibili deployed 1 billion RMB in traffic support for video podcasts, while prominent figures like Joe Rogan and Lu Yu joined long-form dialogue content. Xiaohongshu and Douyin Select are also expanding audio and video capabilities.
A sector once labeled ’niche,’ ‘slow to ignite,’ or even dependent on ‘passion-driven labor’ is now experiencing large-scale commercial growth. Xiaoyuzhou’s commercial revenue surged by 760% in 2025, with nearly fourfold more podcasts achieving monetization.
Crucially, this growth stems not from a few头部s but from consistent updates across multiple programs and diversified revenue paths. Brand partnerships are evolving from one-off content placements toward long-term, program-aligned collaborations. Revenue channels now span brand ads, paid content, and hybrid online-offline initiatives.
C-end paid content GMV grows over 50% annually, with B-end accelerating faster. ‘Vango MoneyTalk,’ for instance, launched recurring annual subscriptions at premium pricing with sustained listener support—proof that trust-built content supports long-term revenue.
Under AI pressure: Human voice and立场 (position) remain irreplaceable
As AI lowers production costs across voice, script, and opinion generation, Xiaoyuzhou has taken protective measures: AI-generated bulk content faces back-end distribution limits, though not yet visible to end users.
CEO Kyth clarifies the core thesis: Podcasts’ true moat lies in ‘authenticity’—consistent立场 from real individuals speaking in their original voice.
The key distinction:
- AI produces zero-cost, unlimited iterations with no名誉 stakes (reputational exposure);
- Human creators cannot reinvent their stance repeatedly without losing credibility.
The ‘奥本海默’ (Oppenheimer) example illustrates this: five podcasts giving scores of 1, 5, 3 respectively—with distinct perspectives—is valuable. But five identical AI episodes with scores 1→2→3→4→5 lacks立场 and wastes listener time.
Voice update features lower expression barriers, encouraging spontaneous, authentic exchanges. Unexpected usage—like in-app karaoke—within two weeks confirms strong demand for low-friction authenticity.
Video and audio: Complementary, not substitutive
Video podcasts raised concerns about podcasts becoming ‘heavier.’ Kyth clarifies the design philosophy: video is optional, on-demand enrichment—not a mandatory viewing requirement.
Evidence comes from real usage: two-thirds of U.S. video podcast consumption occurs as background audio. Xiaoyuzhou aims for seamless switching—‘watch when you want to see, switch back to audio when listening.’ Video adds value particularly for comedy and talk shows, mirroring the full-of-expression live-recording environment.
Ultimately, the move satisfies a deep user need in the fragmented era: connecting with people—not just consumption—through a single speaker’s window into diverse ideas.
The path forward: making podcasts discoverable and defending irreplaceable formats
After achieving initial product-market fit three years ago, Xiaoyuzhou’s next PMF goal is facilitating discovery: helping new users easily grasp and engage with podcasts. Features like transcripts and initial recommendation-feed tests lower entry barriers.
Kyth outlines three main objectives:
- Significantly enhanced content diversity and richness;
- Sharply improved discoverability for creators;
- Expanded and diversified commercial opportunities.
Xiaoyuzhou also commits to preserving formats only podcasts can support effectively. Shows like ‘Sisyphus Express’—launched by print journalists—show how podcasts absorb deep-reporting audiences诉求, defends against superficial content, and provides an uninterrupted space.
Final thoughts
Xiaoyuzhou’s trajectory confirms a truth as platforms mature: when viral traffic fades, the real competition shifts to trust in authenticity. Tools boost efficiency, but human立场 (立场 =立场), voice, and irreproducible consistency remain the uncopyable core of podcast ecosystems.