Live
🔥 OpenBMB/ChatDevGitHub Trending🔥 microsoft/agent-lightningGitHub Trending🔥 apache/supersetGitHub Trending🔥 shanraisshan/claude-code-best-practiceGitHub TrendingA-SelecT: Automatic Timestep Selection for Diffusion Transformer Representation LearningarXivGUIDE: Resolving Domain Bias in GUI Agents through Real-Time Web Video Retrieval and Plug-and-Play AnnotationarXivSommelier: Scalable Open Multi-turn Audio Pre-processing for Full-duplex Speech Language ModelsarXivCANGuard: A Spatio-Temporal CNN-GRU-Attention Hybrid Architecture for Intrusion Detection in In-Vehicle CAN NetworksarXivDesignWeaver: Dimensional Scaffolding for Text-to-Image Product DesignarXivA Lightweight, Transferable, and Self-Adaptive Framework for Intelligent DC Arc-Fault Detection in Photovoltaic SystemsarXivConsistency Amplifies: How Behavioral Variance Shapes Agent AccuracyarXivStabilizing Rubric Integration Training via Decoupled Advantage NormalizationarXivSemi-Automated Knowledge Engineering and Process Mapping for Total Airport ManagementarXivAIRA_2: Overcoming Bottlenecks in AI Research AgentsarXivBeSafe-Bench: Unveiling Behavioral Safety Risks of Situated Agents in Functional EnvironmentsarXiv🔥 OpenBMB/ChatDevGitHub Trending🔥 microsoft/agent-lightningGitHub Trending🔥 apache/supersetGitHub Trending🔥 shanraisshan/claude-code-best-practiceGitHub TrendingA-SelecT: Automatic Timestep Selection for Diffusion Transformer Representation LearningarXivGUIDE: Resolving Domain Bias in GUI Agents through Real-Time Web Video Retrieval and Plug-and-Play AnnotationarXivSommelier: Scalable Open Multi-turn Audio Pre-processing for Full-duplex Speech Language ModelsarXivCANGuard: A Spatio-Temporal CNN-GRU-Attention Hybrid Architecture for Intrusion Detection in In-Vehicle CAN NetworksarXivDesignWeaver: Dimensional Scaffolding for Text-to-Image Product DesignarXivA Lightweight, Transferable, and Self-Adaptive Framework for Intelligent DC Arc-Fault Detection in Photovoltaic SystemsarXivConsistency Amplifies: How Behavioral Variance Shapes Agent AccuracyarXivStabilizing Rubric Integration Training via Decoupled Advantage NormalizationarXivSemi-Automated Knowledge Engineering and Process Mapping for Total Airport ManagementarXivAIRA_2: Overcoming Bottlenecks in AI Research AgentsarXivBeSafe-Bench: Unveiling Behavioral Safety Risks of Situated Agents in Functional EnvironmentsarXiv

Apple’s AI Playlist Playground is bad at music

The Verge AIby Terrence O'BrienMarch 26, 20264 min read0 views
Source Quiz

Apple Music: "What do you want to hear?" Me: "Atmospheric instrumental black metal to write to." Apple Music: "Here's three metal songs with vocals, a field recording, an ambient electronic track, and a piece of doom jazz." I am skeptical of AI's ability to serve up the music I want to begin with, but even […]

Apple Music: “What do you want to hear?”

Me: “Atmospheric instrumental black metal to write to.”

Apple Music: “Here’s three metal songs with vocals, a field recording, an ambient electronic track, and a piece of doom jazz.”

I am skeptical of AI’s ability to serve up the music I want to begin with, but even I was caught slightly off guard by how underwhelming Apple’s new Playlist Playground beta is. YouTube Music’s AI playlist generator is far from perfect, but when I gave it the same prompt for instrumental black metal, it wasn’t until the fifth track that it delivered something with lyrics, and that was the exception, rather than the rule. Apple Music failed to deliver from moment one, and did so repeatedly.

I was in a metal mood yesterday, so I also asked Playlist Playground to create me a playlist from the prompt, “modern ambient black metal from the American South.” Apparently, Apple could only find three songs that would possibly fit that criteria. And one of those was by the band Woman is the Earth from South Dakota. Now, I am an American, and we’re notoriously bad at geography. But, I’m 99.999 percent sure that South Dakota is not the South.

Maybe Apple just isn’t up on black metal. So I tried something I thought would be a bit easier and prompted it for “kid-friendly modern hip hop.” The first track was just the censored version of Kendrick Lamar’s “DNA.” I suppose you could argue that, since it’s the censored version, it technically qualifies.

Next, it served up the censored version of Kid Capri’s “We’re Unified,” which came out in 1998. I am a middle-aged man with a pretty generous definition of what qualifies as “modern,” but this is not it. Six of the 16 songs it added to the playlist were over 15 years old. Three were over 25 years old.

The biggest problem was the inclusion of “ABC” by Chicken P, which is an alphabetical list of all the women he’s slept with and includes lines like, “Desiree, take dick, she be telling me ‘go deeper.’” I’m not sure I need my four-year-old repeating that to their classmates.

The failure here is particularly galling when the Spider-Verse soundtracks, Tyler the Creator’s The Grinch EP, and Aesop Rock exist.

I figured a prompt for “industrial-influenced dance punk” would be simple enough. I was wrong. I was thinking of two specific bands, Model/Actriz and Special Interest. Neither appeared on the playlist. Half of it was just straight old-school industrial: Cabaret Voltaire, Einstürzende Neubauten, Ministry, Front 242, and Nine Inch Nails each made an appearance. But so did New Order’s “Blue Monday” and Irish rappers Kneecap.

Playlist Playground is in beta, and obviously, some issues are to be expected. But based on my results, I’d say it’s not even ready for a public beta yet. Even when it wasn’t failing in spectacular fashion, the results were just kind of dull and rarely served up artists I wasn’t already familiar with. Apple didn’t respond to a request for comment.

Apple’s AI playlist maker clearly struggles with understanding genre, geography, time, and lyrical content. So I gave it one last test and asked it to create a playlist for “School pickup on a chilly day with no children’s music.” I might as well have just put on the Garden State soundtrack (Shins, Nick Drake, José Gonzalez). But, at least I didn’t have to listen to any Parry Gripp.

Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.

  • Terrence O'Brien

Original source

The Verge AI

Was this article helpful?

Sign in to highlight and annotate this article

AI
Ask AI about this article
Powered by AI News Hub · full article context loaded
Ready

Conversation starters

Ask anything about this article…

Daily AI Digest

Get the top 5 AI stories delivered to your inbox every morning.

Knowledge Map

Knowledge Map
TopicsEntitiesSource
Apple’s AI …Music Gener…AppleHallucinati…Generative …The Verge AI

Connected Articles — Knowledge Graph

This article is connected to other articles through shared AI topics and tags.

Knowledge Graph100 articles · 339 connections
Scroll to zoom · drag to pan · click to open

Discussion

Sign in to join the discussion

No comments yet — be the first to share your thoughts!