Spotify vient d'annoncer un truc intéressant je trouve, qui s'appelle les Personal Podcasts. Le principe c'est de demander absolument tout ce que vous voulez, par exemple un podcast sur un cours que vous venez de suivre, sur un bouquin, sur un article de Korben.info voire sur votre planning de la semaine, vos objectifs...etc etc... Peu importe... Spotify prend tout ça, et génère un épisode audio personnalisé raconté par une voix IA plus ou moins moche.

Et l'épisode apparaît alors directement dans votre bibliothèque, comme si quelqu'un avait fait un résumé audio de votre semaine.

Et pour faire ça, ils ont mis en ligne un repo GitHub save-to-spotify qui est un outil en ligne de commande permettant à des agents IA de créer ce podcast personnalisé tout ça accessible en ligne de commande ou via des agents comme Claude Code, OpenClaw ou Codex.

Que ce soit sous macOS ou Linux, ça s'installe en une ligne (allez lire le install.sh par sécurité quand même avant de l'exécuter) :

curl -fsSL https://saveto.spotify.com/install.sh | bash

/plugin marketplace add spotify/save-to-spotify

Design Mindset is Yanko Design’s weekly podcast, powered by KeyShot, the 3D rendering and visualization software that helps designers test how products feel, not just how they look. Hosted by Radhika Seth, the show goes deep into the philosophy and process behind world-class products, sitting down with the designers and founders who actually built them. Episode 19, premiering this week, is one of the most thought-provoking conversations the series has produced yet.

Joseph Hofer is the founder of Hofer Studio, where he consults with hardware entrepreneurs on building profitable, world-class product portfolios. Before that, he spent over a decade at BlackBerry as senior industrial designer, establishing the look and feel of the iconic Bold family and shaping devices like the Q10, Z10, and the BlackBerry Passport. His work spans over sixty design and utility patents, touching products that have sold over twenty-one million units and generated upward of $3.1 billion in revenue. More recently, he’s been the design force behind the Clicks Communicator, a physical-keyboard phone that launched at CES and challenges the smartphone status quo from the ground up.

Download your Free Trial of KeyShot Here

Designing Within Human Limits and Intentional Use

Joseph opens the conversation with something that sounds almost poetic but lands with the weight of a core design principle, saying that “most of the objects we use every day quietly train us. They teach us how to hold them, how long to focus, how patient we need to be. When design ignores human limits, it drains us. When design respects them, it almost feels like care.” He critiques what he calls “sticky” experiences, the kind that benefit companies at users’ expense, arguing that the real question designers should be asking is whether a product helps people become a better version of themselves, or whether the company simply wins after ten years of draining them.

His case against the modern smartphone is pointed. Everything phones have become reactionary devices, he says, describing the experience of opening one to send an email and somehow finding yourself fifteen minutes deep in a reel, asking yourself how you got there. Big tech, in his view, has deliberately shaped products to increase screen time and sell more through ads. His philosophy runs in the opposite direction: good design should prompt intention before action, not exploit the absence of it.

Integration as Core Design Principle

One of the more revealing details Joseph shares early on is that at BlackBerry, the design team’s official title wasn’t “Industrial Design.” It was Design Integration. That framing stayed with him. “Integration is probably the word, the action that I look to do well in every project I work on,” he says, adding that a product can be really strong in one area but fall flat in others if you’re only focused on a single dimension. Great design, strong UX, and poor profit economics don’t add up to a sustainable company. Economics, manufacturing, cost, and complexity all have to be part of the thinking from the start.

His advice to technical founders reflects the same logic. Many of them start with a breakthrough innovation and then go looking for a market to push it into, which he sees as working in the wrong direction. The better path is to step back, clearly analyze the problem bubbling up from the market, shape an experience that solves it, and then let the technology marry with that. Letting one run too far ahead of the other is how good innovations end up as products nobody uses.

The Clicks Communicator: Intentional Mobile Interaction

The Clicks Communicator is the most direct physical expression of Hofer’s philosophy. It was the first phone he designed in ten years after BlackBerry, and the central idea is a complete inversion of how smartphones currently work. Rather than an app grid that presents notifications and pulls users in reactively, the Communicator prompts users to decide what they want to do first, then acts on it. Physical keys map to intentional shortcuts: pressing K calls a specific contact, pressing I opens Instagram only when the user has consciously chosen to. “It flips it from being reactionary to intentional,” Joseph says simply.

He’s also clear that the product’s appeal isn’t nostalgia. A lot of the customers aren’t even BlackBerry users, he notes; they’re younger people who simply want a different relationship with their mobile device. The Communicator sits within what he sees as a broader 2025 trend of “intentional tech,” products designed to decouple from the everything-phone model and serve one specific purpose well. Adding a 3.5mm headphone jack and a removable SD card wasn’t feature-stacking for its own sake either; those choices are signals to a specific audience that the team is listening and cares about them.

Recognizing Quiet Ideas and Process Discipline

When Radhika points out that the BlackBerry keyboard now feels like it was always inevitable, Hofer pushes back immediately. “Sometimes these quiet ideas that feel obvious or become obvious actually took a lot of effort and iteration to get there,” he says, describing the motto his team lived by: think, build, test. The keyboard’s evolution wasn’t a single stroke of insight; it was a response to real constraints. As iPhones pushed screens larger, BlackBerry faced intense pressure to shrink keypads, which meant switching from oval keys to square ones, losing the tactile separation users relied on. The innovation was subtle: raising a curved edge on each square key to preserve the feeling of the oval, essentially hiding a reference to the old shape inside the new form. Speed tests, accuracy tests, user sentiment on different options, all of that grinding iteration is what produced something that feels natural.

He applies the same thinking to simplicity broadly. Designing for a ten-year-old, he argues, is one of the most useful principles any designer or founder can adopt. If you can’t explain the product to a ten-year-old, it’s too complicated. He tested this literally the night before the recording, sitting down with his eight-year-old daughter to ask about her CD player. Her answer was that it had way too many buttons. Her ideal? Three: power, volume up, volume down. Six identical-feeling buttons with in-mold graphics that disappear in the dark told a clear story about what the designers had gotten wrong.

Restraint as Confidence and Commercial Strategy

The tension between restraint and visibility is something Hofer takes seriously. He doesn’t frame minimalism as a virtue in itself. “Clarity is actually an even stronger word,” he says, arguing that a vanilla product solving a vanilla problem will simply go unnoticed. The goal isn’t to be quiet; it’s to solve a real, specific problem so well that the product becomes the only answer for a particular group of people. A phrase he came up with captures where he’s trying to take the companies he works with: from viral products to vital ones, products that customers genuinely need in their lives because of the difference they’ve made.

That philosophy maps directly onto commercial outcomes. A product that meets the emotional and functional needs of a user, reduces cognitive load, lasts longer, and has lower return rates naturally builds a brand that draws people in without needing to be aggressively sold. “When products are just better,” he notes, “they need to be marketed and sold maybe less. That’s an effect on your bottom line.” His work at Hofer Studio is less about crafting beautiful objects and more about asking founders what commercial success actually means to them and building backwards from that.

When the rapid-fire round asks him to describe restraint in design in a single word, his answer arrives without hesitation: confidence. “What does obviousness create? It creates confidence. I know how to enter this experience. I know how to start this product. I feel more confident with it in my life.” It’s a fitting close to a conversation that consistently returned to the same idea: that the design decisions nobody notices are usually the ones that took the most care to make.

Design Mindset drops every week on Yanko Design. Catch Episode 19 in full wherever you listen to podcasts. For a free trial of KeyShot, visit keyshot.com/mindset.

Download your Free Trial of KeyShot Here

The post Ex-BlackBerry Designer Is Calling Out Everything That’s Wrong With Modern Phones first appeared on Yanko Design.

Vous vous souvenez d’Anchor, cette plateforme qui avait permis de simplifier la création de podcasts avant de se faire racheter par Spotify pour 150 millions de dollars ? Eh bien ses créateurs, Nir Zicherman et Michael Mignano, sont de retour avec un nouveau projet qui pourrait bien changer l’apprentissage en ligne. Leur nouvelle création s’appelle Oboe et permet de créer des cours éducatifs personnalisés en quelques secondes grâce à l’IA.

Une fois sur Oboe, vous tapez un prompt du genre “Histoire de l’IA”, “Comment réaliser des pâtisseries” ou même “Préparer mon déménagement au Japon” et hop, Oboe vous génère instantanément un cours complet adapté à vos besoins. C’est le premier outil d’apprentissage généralisé par IA qui permet vraiment de créer des formations sur n’importe quel sujet, d’une manière plus personnalisée et efficace que tout ce qui existe aujourd’hui.

Mon cours sur les échecs

Ce qui rend Oboe vraiment intéressant, c’est son architecture multi-agents développée from scratch. Sur TechCrunch , Zicherman explique que leur système fait tourner plusieurs agents IA en parallèle qui s’occupent de tout : développer l’architecture du cours, vérifier les informations de base, écrire le script pour les podcasts, récupérer des images pertinentes sur Internet et même auditer le contenu pour garantir sa qualité et son exactitude. Tout ça en quelques secondes chrono.

Au lancement, Oboe propose déjà 9 formats de cours différents ( Cours classique, épisode de podcast, points clés à retenir, livre audio, FAQ, flashcards, quiz, etc…) capables de couvrir des domaines aussi variés que les sciences, l’histoire, les langues étrangères, l’actualité, la pop culture ou même la préparation aux changements de vie selon ce que vous lui demandez. Chaque cours combine du texte, des visuels, des conférences audio, des quiz interactifs et même des jeux. Vous avez deux modes audio disponibles : un format de conférence traditionnelle ou une expérience plus conversationnelle façon podcast, un peu comme NotebookLM de Google.

Le nom “Oboe” n’a pas été choisi au hasard. Il s’inspire de la racine du mot japonais qui signifie “apprendre”. Une jolie référence qui colle parfaitement à leur mission de rendre l’humanité plus intelligente. Encore faudrait-il qu’elle souhaite “apprendre”, cette humanité…

Niveau business model, Oboe adopte une approche freemium assez classique. Vous pouvez créer jusqu’à 2 cours par mois et si vous voulez aller plus loin, deux forfaits payants sont disponibles : Oboe Plus à 12$/mois pour 30 cours supplémentaires et Oboe Pro à 32$/mois pour 100 cours.

Au lieu d’avoir des cours statiques créés par des experts et consommés passivement, Oboe permet à chacun de générer instantanément du contenu éducatif ultra-personnalisé, comme un prof particulier capable de créer un programme sur mesure en temps réel, quel que soit le sujet qui vous intéresse.

Perso, j’ai testé et j’ai trouvé les 2 formations que ça m’a généré de très bonne qualité. Et si vous creusez un peu sur Google, vous pouvez aussi trouver certains cours générés par d’autres utilisateurs : inurl:https://oboe.fyi/courses/

Quand on voit la qualité de ce que peuvent générer les IA aujourd’hui, l’idée d’avoir une plateforme capable de créer des cours complets et structurés en quelques secondes, c’est assez cool, surtout avec leur système multi-agents qui vérifie la qualité et l’exactitude des informations. Ça pourrait vraiment changer les choses dont on apprend de nouvelles compétences ou dont on explore de nouveaux sujets.

Reste maintenant à voir si Oboe réussira à s’imposer face aux géants de l’éducation en ligne mais avec l’expérience des fondateurs, ils ont clairement quelques atouts dans leur manche.

Il y a quelques semaines, j’ai eu l’occasion de participer au podcast ‘Derrière La Bulle’ de Léo Pereira et c’était vraiment sympa !

On a fait un voyage dans le temps, depuis l’époque où uploader une image de 50 Ko prenait une pause café jusqu’aux IA génératives du moment et comme j’ai vécu toute cette évolution des débuts chaotiques du HTML, à l’émergence des blogs, en passant par la montée du web social, et maintenant l’ère de l’IA, j’avais plein de trucs à raconter.

Si vous passez vos journées, le nez dans un terminal, ce serait-y pas trop cool de pouvoir également y écouter vos livres audio et vos podcasts ?

On évite ainsi les interfaces flashy, les notifs et autres animations inutiles et on se concentre sur l’essentiel, à savoir la simplicité monacale du terminal. Bref, si vous êtes de ceux qui pensent qu’une journée sans taper une commande est une journée perdue, il faut que je vous présente Toutui !