Machine Learning Engineering Manager - Music

HYPHNT footwear, Pan Macmillan and Spotify unveil world's first audiobook trainers. HYPHNT footwear, publisher Pan Macmillan and the world's most popular audio service Spotify have joined forces to launch the world's first audiobook trainers The collaboration seamlessly merges fashion with storytelling for the National Year of Reading. The £79 HYPHNT footwear 'Nova' junior trainer features embedded Near Field Communication (NFC) technology that connects wearers directly to audiobooks on Spotify, debuting with John Patrick Green's bestselling InvestiGators graphic novel series. A simple tap of a smartphone against the shoe instantly launches the Spotify app, where they can discover the featured audiobook. New books and stories from Pan Macmillan will be introduced over time, giving young wearers a reason to return, tap again and discover what's next. The launch addresses shifting youth literacy habits, backed by data showing audio's growing role in how children learn and explore. Nearly half (47%) of children and young people listen to audiobooks, podcasts, or both, and 56% of these listeners tune in at least once a week. Half (50%) of young audiobook listeners state that listening "feels like reading," while 37% say audiobooks actively sparked their interest in reading more broadly. Meanwhile Spotify's research shows that 68% of parents who listen to audiobooks use them to help maintain a strong reading culture at home. Don't Miss Grab your pair! Jason Black, founder of HYPHNT footwear, said: 'Pan Macmillan has always believed a trainer can be so much more than a physical product. Having used embedded NFC tech in its adult range to unlock music and gaming, bringing connected footwear to young readers felt like a natural evolution. Kids don't experience culture in rigid silos - they love trainers, tech, music, and books all at once. Partnering with Pan Macmillan and Spotify allows Pan Macmillan to bring those worlds together seamlessly. InvestiGators is just the starting point; kids can tap the same pair of shoes over time to unlock new audiobooks, making reading a high-status lifestyle statement on the pavement.' Jodie Williams, Head of DEI and Social Impact at Pan Macmillan, said: 'The National Year of Reading is about finding new ways into stories, especially for those who don't yet see themselves as readers. Libraries and classrooms will always be central to that, but reading should also pop up in the everyday moments of its lives. Embedding audiobooks into footwear sounds playful, and it is, but the idea behind it is simple: when books turn up in unexpected places like a pair of trainers, it shows that reading belongs anywhere.' Jamie Forrest, Communications Director, Macmillan Children's Books added: 'Bringing Mango and Brash off the page and into footwear turns reading into an active, everyday adventure. InvestiGators has captured millions of young imaginations, and giving kids an interactive way to step directly into their favorite series makes discovering a great book fun and effortless.' Roshni Radia, Books Editor at Spotify said: 'Pan Macmillan is always looking for new ways to help people discover great stories, and Pan Macmillan love the idea of giving young people quick and easy access to audiobooks in an unexpected part of their everyday lives. Its collaboration with HYPHNT and Pan Macmillan makes discovering your next listen as simple as tapping your trainers, and is a fun way to get more young people excited about books and storytelling.'

Music and sound generation powered by ElevenLabs and Stability AI. Spotify just handed an AI agent the aux cord - search bars may be next. The next person choosing your music may not be a person - and it may not live inside your streaming app. Spotify has connected to Meta's Muse personal AI agent, letting listeners play audio, save songs and build playlists through a conversation. Spotify is the first music streaming service connected to Muse, according to coverage published September 24. This sounds like a convenient voice shortcut. It is actually a preview of a much larger change: the layer between listeners and music is becoming an agent. What Muse can do with Spotify. After connecting their accounts, listeners can ask Muse to: * play music, podcasts or audiobooks; * save songs to their Spotify Library; * create and edit playlists; * find something based on a conversational request; * resume audio on a connected device; * turn notes or briefings into a Personal Podcast saved to Spotify. The interesting part is context across services. In Spotify's example, Muse can notice a road trip on a calendar, build a playlist inspired by the destination and schedule it to begin when the trip starts. That is not "play my favorites." It is an agent combining time, place, intent and listening history to make a programming decision. Streaming platforms have used machine learning for recommendations for years. The old model asked: "People with behavior like yours played this - might you like it?" An agent can answer a more specific request: Build a two-hour driving playlist for leaving Los Angeles at sunset, keep the first 20 minutes calm, raise the energy after the freeway clears, avoid songs I heard this week, and save three discoveries for later. The listener no longer has to translate that intention into searches, filters and manual queue edits. The agent does the translation - and chooses which metadata, sonic traits and behavioral signals matter. Spotify has been moving in this direction through several doors. It launched a ChatGPT app integration in October 2025 and a Claude connector in April 2026. Its own interface has also become more conversational. Muse is the clearest sign yet that music control is escaping the music app itself. The new gatekeeper is the request interpreter. For artists, this changes discovery. Today's optimization advice revolves around playlist editors, follower conversion, short-form clips and metadata. Agent-mediated listening adds another question: can a system understand when your track fits a richly described moment? That could reward music with accurate metadata and a clear sonic identity. It could also concentrate power in opaque recommendation systems that decide what "late-night focus," "first-date confidence" or "rainy train ride" sounds like. The challenge is not stuffing descriptions with keywords. Agents can analyze audio and behavior. The durable advantage is making a track that has a legible use, mood and audience - then describing it honestly. What musicians should do now. * Treat metadata as creative infrastructure. Genre, mood, tempo, instrumentation, credits and clean release information help systems place music correctly. * Build distinct versions intentionally. Instrumental, clean, extended and stem-based variations can serve different agent requests. * Name the moment your track serves. "Electronic" is weak. "Patient, low-vocal electronic music for focused night work" is useful. * Protect your identity signals. As agents learn taste, consistent artist pages, credits and catalogs matter more, not less. * Do not confuse volume with discoverability. Flooding services with near-identical tracks makes it harder for people - and eventually agents - to understand why any one track matters. The search bar is not disappearing tomorrow. Muse is new, and 2.8 million reported downloads in its first 12 days across the U.S. and Canada do not make it the default way people listen. Spotify's native app, playlists and search remain dominant. But the direction is visible. Music is becoming one capability inside broader personal agents that know the calendar, the destination, the workout and the task. When that happens, the winning track may not be the one a listener searches for. It may be the one an agent decides is right before the listener opens the app. The aux cord did not disappear. It learned how to talk. Sources.

Spotify now lets listeners tell its algorithm what they want. Spotify's Taste Profile gives eligible Premium listeners a way to describe what they want to hear next. Here's how to use it, and why this matters. Sep 23, 2026 Spotify is rolling out Taste Profile (Beta) to Premium listeners 18 and older in the US, following a test in New Zealand. The feature lets people see a short account of how Spotify understands their listening across music, podcasts and audiobooks, then tell the app what they would like more or less of on their Home feed. Usually, listeners shape recommendations through what they play, skip and save. Taste Profile gives them a way to state an interest before their listening history reflects it: they can ask for more of a genre they have just discovered, music that suits a particular part of the day, or less of something they have moved on from. These requests guide recommendations; they do not issue commands. Spotify says Taste Profile cannot follow an artist on someone's behalf or enforce rules such as "always show" or "never show." How to use Spotify Taste Profile. Eligible listeners can open the Spotify mobile app and: * Tap their profile picture, then select Taste Profile. * Read Spotify's summary of their interests across music, podcasts and audiobooks. * Under Tell Hypebot more, type what they want to change and send the message. Someone might ask for more contemporary jazz, more music for an early morning run, or fewer recommendations tied to a passing obsession. Spotify says changes typically appear within a few hours. Listeners can edit or delete their notes, or leave their profile alone if they like what Home already shows them. The beta is available to eligible Premium listeners in the US and New Zealand; Spotify has not announced a broader rollout in its September 23 post. What could change for artists? For independent musicians, the interesting possibility is that a listener can now express curiosity before they have streamed enough music to demonstrate it. If someone tells Spotify they want to explore a particular sound, relevant artists could have another opportunity to appear in that listener's Home recommendations. Personalized recommendations already play a substantial role in discovery. Spotify has said that 35% of discoveries on its platform in 2024 happened through personalized recommendations in algorithmic contexts. Taste Profile adds a listener's stated interests to the picture, but Spotify has not described a way for artists to target those notes or see which request led a listener to their music.