New Google ‘mind-reading’ AI can predict what music you listened to

Hotstar in UAE
Hotstar in UAE

In recent years, the development in the AI field has truly been astounding, with generative AIs like ChatGPT quickly becoming a part of our lives. Now, as part of these efforts, scientists at Google have achieved a significant breakthrough in ‘mind-reading’ AI, allowing them to decode individuals’ music preferences by analyzing their brain signals.

Dubbed Brain2Music, the system utilizes functional magnetic resonance imaging (fMRI) data, which tracks the blood flow to different brain regions, and brain-computer interfaces (BCIs) to capture neural activity patterns as listeners hear different songs. According to Timo Denk, a co-author of the study and a software engineer at Google in Switzerland, the AI successfully predicted the agreement between the mood of the reconstructed music and the original music about 60% of the time.

“The method is pretty robust across the five subjects we evaluated. If you take a new person and train a model for them, it’s likely that it will also work well,” said Denk.

How does Brain2Music work?

Google says it first trains the AI using a subset of brain imaging data and accompanying song clips with the objective of establishing connections between various musical elements—such as instruments, genre, rhythm, and mood—and the participants’ brain signals. Then, as participants listen to music, the AI system meticulously analyzes neural patterns and cross-references them with music metadata.

Over time, the machine learning models become adept at recognizing distinct patterns associated with specific musical elements, such as rhythm, melody, and harmony, and subsequently associate them with corresponding songs in the database.

 

Ethical complications

While this breakthrough represents a significant development for Google in the AI field and could have immense potential for various neurotechnology applications, the mind-reading capabilities of the AI have once again prompted questions about privacy, consent, and data usage. Therefore, if Google plans to commercialize this technology in the future, it would need to implement stringent data protection measures to ensure user information remains confidential.

2023-08-07 15:05:44