A couple of days ago, Soul App launched its latest AI-powered function that is pegged to change social play. Named “Lazy KTV”, this is a music-centric function that is excitingly different from anything out there in the social space at this time.

Before the launch of “Lazy KTV”, Soul had launched its predecessor, the “Soul Dimension Singer” in 2023. At its introduction, this function too had built user anticipation to a crescendo as it allowed them to create their very own AI singer avatar.

Both these functions are a result of Soul App’s investments in the use of speech and music for social interactions and engagement. Music is an important "language" of communication that can bring together people from all walks of life. And the app has emphasized interest-based social interaction from the get-go. So, it is understandable how and why music has turned into one of the most popular interest categories on the platform.

In fact, the music-centric functions of Soul App have managed to grab the attention of diverse audiences. From high-quality music creators to those simply interested in music creation for the fun of it and even casual users who are seeking spontaneous relaxation and entertainment through music comments and music and lyric sharing, Soul App has managed to impress them all with its suite of music-related functions.

Soul’s foray into the world of music as a form of interaction and connection started with the introduction of the Linglun engine, a proprietary creation that boasted industry-leading audio deep-learning capability. The introduction of Linglun Version 1.0 opened up the doorway for the eventual roll-out of the “Soul Dimension Singer”.

Given the positive user response to the platform’s first exclusive music-centric function, the team behind Soul App quickly got to work on Linglun Version 2.0, which was launched a few months ago. The latest version of the music creation engine is better than its predecessor in every aspect.

For example, in terms of vocal range control, it has been upgraded to a multi-person multi-scale self-adaptive model. This markedly improves the similarity and expressiveness control of multi-person synthesis. As far as singing voice synthesis goes, Linglun 2.0 has been upgraded to the cutting-edge denoising diffusion probability model. This gives it the ability to offer enhanced quality and accuracy of song and voice synthesis.

These improvements in Version 2.0 made it possible to build an innovative function like “Lazy KTV” that offers a multi-person UGC "AI Duet" experience. Using the “Lazy KTV” function is simple. Users start by recording a 30-second audio clip, which serves as a personalized voice model. This clip is then put through one of the three AI singing modes - balanced, similarity, and expressiveness in preparation for further synthesis. The result of this process is a personalized and unique music cover.

But, the “Lazy KTV” feature is not just about solo singing. Its most notable aspect is that it also supports the "AI Duet" mode. So, users can invite friends to complete voice cloning together and perform a musical "duet" across time and space. In fact, it is this differentiated interactive experience that has users the most excited.

Within days of its launch, Soul users have actively participated and invited friends to experience the "AI Duet" mode of the function. And why not? Singing with friends is by far one of the most fun and engaging activities; not to mention that it is also incredibly effective at chasing away stress and fatigue and helping people unwind. Plus, it is a great way to engage with friends and acquaintances at a whole new level.

From here, the next step for the Soul App team in this musical journey is to further refine the features of “Lazy KTV”. Word is that they are already working on a one-click music creation feature based on AIGC technology that will allow users to churn out personalized interactive musical creations, with the right lyrics, melody, arrangement, singing, and mixing.

This feature is expected to further lower the threshold for users to express themselves through music, enjoy themselves with music, resonate with music, and socialize with music. It is also expected to meet the needs of young people for a more immersive and differentiated interactive experience.

While music is currently at the forefront of Soul App’s “AIGC + Social” strategy, this certainly isn’t the first time that emerging and innovative technologies have been used to power unique features on the platform. Before music turned into a focal point, the Soul App team was using AI to support intelligent dialogue and image creation.

As a social platform that has extensively used AI technology to solve the needs of human connection, Soul has successively launched the self-developed:

1. Lingxi Engine (an intelligent recommendation system based on AI technology).
2. NAWA Engine (an integrated SDK that enables AI avatar processing, AR augmented reality, PTA image generation, STA speech generation animation, TTA text generation animation, and other technologies).
3. Soul X (a self-developed vertical language model that is the basis of the intelligent dialogue robot "AI Goudan", AI-assisted chat, virtual companionship, and other diverse functions).

The aim all along has been to bring users a more immersive, intelligent, and in-depth social experience on the platform. And given this trend, it’s obvious that Soul App will continue to use AIGC technology to introduce other unique functions that will cater to the ever-changing demands of platform users.