We Are Social Asia’s Tuesday Tune Up


Meta open sources an AI-powered music generator

Meta has released its own AI-powered music generator called MusicGen. The music-generating tool can turn a text description (e.g. “A ’90s pop song with synth pads in the background”) into about 12 seconds of audio. MusicGen can optionally be inspired by reference audio, i.e an existing song, in which case it’ll try to follow both the description and melody. MusicGen was trained on 20,000 hours of music, including 10,000 “high-quality” licensed music tracks and 390,000 instrument-only tracks from ShutterStock and Pond5, a large stock media library. 

You can now share music on Instagram Notes, and view translations

Following the public announcement of new features to Instagram Notes back in December last year, Notes will now support music and translations, announced Meta CEO Mark Zuckerberg via his Instagram Broadcast Channel. Instagram Notes allows users to leave a status message that appears for their Close Friends or mutuals at the top of their inbox, and until now, could only include text and emoji. The new feature will allow users to share a clip of a song and a caption to the Note if they like. Song clips will be supported in all markets worldwide where Instagram has music licensing rights.

Instagram Notes will now also have a ‘translations’ feature for posts that are in a different language, as is available on ​​Post descriptions, comments, and Reels.

Instagram is rolling out its Channels feature for broadcasting messages globally

Meta is rolling out Instagram’s Channels feature for broadcasting messages globally, which has only been available to select creators and users until now. CEO Mark Zuckerberg made the announcement on his Instagram channel, recommending some channels for people to follow. The Channels feature allows creators to share one-to-many messages with their fans, and in May, the company announced the ability for creators to collaborate on channels. 

Meta showcases new ‘Voicebox’ speech-to-text translation tool

Meta has announced Voicebox – a generative AI model that can help with audio editing, sampling and styling. The system can take text inputs and translate them into audio, with different voice options, enabling more advanced text-to-audio translation, but with reduced learning and processing requirements than other, similar offerings. Additionally, Voicebox is multilingual, so will allow simplified, native-sounding variations of text inputs in different languages.

Elon Musk says Twitter video app for smart TVs is ‘coming’
A Twitter video app for smart TVs is in the social media company’s plans, owner Elon Musk said in a tweet on Saturday – a move that would be in line with plans to focus on video creator and commerce partnerships, which CEO Linda Yaccarino and Musk laid out in an investor presentation last week.

In other news

TikTok now allows users to add a 4000-characters description. Instagram is working on the ability to generate comments with AI 👀 On Instagram, you can control GIF comments by adjusting the settings, and TikTok has added a new “On this day” feature to repost videos as a story or new video 📹