Devpost’s Post

At ÜberStrategist, we're not just confined to one domain – we're passionate about all things tech and AI, too! Listen to the latest episode of Tech It Out with Marc Saltzman, which delves into the fascinating work of our client, the Be Fake app. What does authenticity look like in the digital age? https://lnkd.in/ddieawPB #BeFake #AI #TechItOut

All programming on Spotify (Yes, that includes editorial and personalized playlisting) uses both AI and a human touch in curation. In the link below is a great article on playlists drive discovery and help to create your next superfan. #MusicBusiness #MusicIndustry #Spotify #DSPs #MusicDiscovery #AI #Playlists

Data, ML, AI in Music https://lnkd.in/e-auYC2m

🔮 Insightful 🎙️ episode from two of the best at Digitas! Noel Childs and Adam Buhler chatting about context switching, the ever evolving ✨ AI ✨ space, managing change and their desires for continuous learning. Check it out on Spotify now! 🎧 #A76designpodcast #Work #Unicorns #Digitas #AI

I’m always fascinated to see how companies harness AI for different use cases. I love Spotify, it helps me build great music lists that always intrigue me of how they can efficiently personalize millions of lists for their subscribers. Recently, The Wall Street Journal had the same question and provided an inside look at how the platform uses machine learning technology to generate personalized recommendations and how they are starting to employ Generative AI to further enhance the subscriber experience. It’s a fascinating video and well worth the 7+ minute watch. As you watch, if you are like me in the business of helping companies provide the best possible customer experience with personalization, think about how applicable some of the same approaches can be (maybe not the danceability parameter). https://lnkd.in/gUHC5qd6 #AI #machinelearning #cx

So much talk on genAI is focused on models, but I'm far more interested in how genAI products are designed. Building with genAI results in so many interesting design discussions because when genAI models can say anything and sound like anything, the designer's role is to figure out the boundaries of the model and shape it into a useful experience. To get there, that means designers must lead discussions about the expected behaviors of the model and the expected behaviors of the people who will interact with the model. They also imagine and define all the unexpected behaviors, because users will always surprise you. All of these discussions ensure the result is useful for customers, and if you're lucky, enjoyable. Listen to this quick talk on how Spotify's product team thought through the challenges of building a personalized music listening experience. Spotify's had to make design decisions around: - How transparent should they be around the use of AI? Do customers care they're using AI? - What should the AI DJ say? (you can't just leave it up to the model to decide) - How the DJ should sound? (And what voice should it use?) https://lnkd.in/gUVGhyhc #productdesign #uxdesign

Sony Music Group accused AI developers of using its intellectual property to train models without permission. Sony issued warnings to over 700 AI developers and streaming services, including Google and Microsoft, demanding disclosure of the works used and requesting changes to terms of service to prevent data collection. Learn more in #TheBatch: https://hubs.la/Q02yHWHM0

Piotr Dąbkowski and Mateusz Staniszewski, who cofounded ElevenLabs labs which has raised over 101 million US dollars to position themselves as the global leader in voice AI research and product deployment Today, ElevenLabs is perhaps best known for its browser-based speech generation app that can create lifelike voices with adjustable toggles for intonation, emotion, cadence and other key vocal characteristics. Now they are coming for musicians! But this time, it's different. 11 Labs' new music generator looks like it can churn out hit songs FASTER than human producers. Here's a sample: 11 Labs creates insanely realistic, Spotify-ready tracks. But, unlike other AI music like SUNO & UDIO, it requires way less human input. Let us know your thoughts and in the comments & follow for more. #ai #technology #founders #music #software #future

My last post wondered aloud whether autonomous agents could simulate emotions in order to make more human-like decisions. Yup. We can probably do that. "Inspired from Damasio’s theories we developed a mechanism in an artificial agent that enables it to represent emotional states and to exploit them for biasing its decisions... The results of this experiment are analysed and discussed highlighting the demonstrated efficiency of the implemented somatic marker mechanism." Now my question is whether they could simulate MY emotional responses. How could you quickly gather that data to train an agent to respond as you do and would you want to? And would the resulting recommendations be better or worse than the collaborative filtering methods that, say, Spotify employs. https://lnkd.in/gABuDqc8

I loved this episode of The Ezra Klein Show "How to Discover your own Taste" Favourite quote (on the world of AI) "And I have this intuition that in a world where everybody has access to these models and what they’re going to do is create a lowest common denominator of aggregated taste, having a very specific sense of taste, having a very specific sense of discernment becomes more important. And it becomes a real mark of distinction because it’s going to be so easy for people to skip that step in becoming their own human being." https://lnkd.in/gKkMhnPS