Overview of AI Audio Updates: Spotify, Particle, ElevenLabs
Host (per transcript): Jaden Schaefer
This episode surveys recent advances in AI-driven audio across three areas: Spotify’s new prompted playlists, Particle’s podcast-clipping news product, and ElevenLabs’ work on multilingual dubbing and the state of AI audio in publishing. The host also announces a major redesign of AIbox.ai (playground for 50+ models, new pricing tiers) and frames the broader trend: AI is changing how we discover, consume, and monetize music, podcasts, and news.
Spotify: Prompted Playlists
- What it is:
- AI-powered playlists built from a user prompt describing vibe, era, genre, source preference (new vs. library), scenario, etc.
- Example prompt: ask Spotify to find an artist you haven’t listened to, introduce you via top songs, and order likely favorites at the top.
- Availability:
- Rolling out to premium subscribers. Initially tested in New Zealand; launched in UK, Ireland, Australia, Sweden; expanding to US & Canada.
- Pros:
- Highly customizable, can surface new artists and songs you’d otherwise miss.
- Lets users shape their “algorithm” with detailed prompts (vibe, era, movie-score analogies, mostly-new tracks, etc.).
- Cons / limits:
- Not automatically updating — you must re-prompt to refresh content.
- Prompting can be extra friction (users may need saved prompts or “prompt engineering”).
- Beta caps reported (~20–30 prompts) — likely cost/learning-control measures.
- Threat to manual playlist curators/curation economy (less discoverability for human-made playlists).
- Host perspective:
- Useful and probably the future of discovery, but could create more work for users who want dynamic playlists without repeated prompts.
Particle: Podcast Clips & AI News
- Product focus:
- Particle is an AI news app (built by former Twitter engineers) that treats podcasts as first-class news sources.
- New “Podcast Clips” feature finds short, interesting moments from podcasts and inserts them into a news feed alongside related stories.
- Why it matters:
- Saves time: extract highlights instead of listening to full episodes (useful for lengthy shows or time-constrained listeners).
- Allows publishers and listeners to surface unique reporting/insights only available in podcast conversations.
- Business model / pricing:
- Freemium product with Particle Plus: $2.99/month (around $30/year) gives summary controls, multiple voices for audio feed, and premium features.
- Use-case implications:
- Great for listeners who want fast takeaways.
- For podcasters: high-value snippets may increase reach but could change how episodes are produced (more emphasis on standout moments).
ElevenLabs: Dubbing, Voice Cloning & the State of AI Audio
- Capabilities discussed:
- High-quality voice cloning and multilingual dubbing for podcasts and videos.
- Handling multi-speaker identification, voice separation, translation, and timing adjustments (stretching/compression) to maintain lip-sync and natural cadence in dubbed audio.
- Applications:
- Podcast translations into multiple languages (French, Spanish, German, Mandarin, etc.) while preserving the host’s “voice.”
- Monetization & reach: localized audio attracts regional advertisers and increases accessibility.
- Business notes:
- ElevenLabs is powerful but can be costly — the host reports extensive personal spend (order-of-magnitude thousands).
- Dubbing and translation are major revenue opportunities for the company.
- Research / report:
- ElevenLabs produced a report on AI audio in publishing and news (state, trends, implications).
Wider implications & trends
- Media formats converging: articles → audio; audio → clips/dubs; music discovery → conversational prompts.
- Emerging experiences:
- Conversational/interruptible voice interactions.
- Interactive news: news as a back-and-forth conversational experience rather than static articles.
- Multilingual audio on demand — publishers moving to “audio-first” strategies to reach global audiences.
- Key strategic insight:
- Winning won’t be just about the best model, but about the best use cases and product experiences built on AI audio tech.
Notable quotes / takeaways from the episode
- “You shouldn’t have to listen to an entire hour-long episode just to find 45 seconds that actually matter.”
- “It’s not just about who has the best AI model… it’s who can come up with the best use cases.”
- “AI is showing up everywhere all at once — it doesn’t just make things more convenient, it changes how we consume media.”
Actionable next steps / recommendations
- For listeners:
- Try Spotify’s prompted playlists if you’re a premium user and experiment with detailed prompts for custom discovery.
- Use Particle (or similar apps) to get podcast highlights if you want fast news/podcast consumption.
- For podcasters/publishers:
- Consider clipping key moments for distribution and discoverability.
- Explore dubbing and voice-cloning services to reach non-English markets and unlock localization revenue.
- For creators/curators:
- Keep saved prompts/templates for Spotify use to reduce friction.
- Reevaluate playlist strategy: AI-curated playlists may affect follower dynamics and promotion tactics.
- For product teams:
- Focus on UX for reducing prompt friction (preset prompts, “one-click” regeneration, auto-refresh).
- Build compelling use cases (interactive audio, conversational news) rather than optimizing only for raw model quality.
Resources & product mentions
- AIbox.ai — host’s platform (playground for 50+ models, new redesign, pricing tiers down to $8.99 and annual plans with 20% discount).
- Spotify — prompted playlists (premium; rolling availability).
- Particle — podcast clipping, Particle Plus ($2.99/mo).
- ElevenLabs — voice cloning, dubbing, translation tools, and an AI audio industry report.
If you want the core points without listening: Spotify = custom, prompt-driven playlists; Particle = highlight-driven podcast news; ElevenLabs = scalable multilingual dubbing and voice cloning — together they point to a future of conversational, localized, and clip-first audio experiences.