SiriusXM is shuttering its Stitcher podcasting app and merging podcast delivery into its flagship SiriusXM subscription offerings. As of August 29, “the Stitcher app and web listening experience will be disabled,” the company told users this week. Stitcher offered listeners the choice of free-to-listen ad-supported programs or a la carte show subscriptions. It also had the $4.95 per month ($34.99 per year) Stitcher Premium, providing a wide variety of ad-free podcasts. “Subscribers can listen to podcasts within the SiriusXM app and will see an all-new listening experience later this year,” the company said. Continue reading SiriusXM to Close Its Stitcher Podcast App and Site in August

The AI Hub server on Discord has drawn attention from the Recording Industry Association of America, which sent a DMCA takedown notice and is alleging copyright infringement. The users are said to share a wide range of AI voice models, including some based on recognizable performers. Those that may sound familiar are in the style of Stevie Wonder, Frank Sinatra, Rihanna and Bruno Mars. AI Hub reportedly has more than 142,000 members that engage in sharing topical information, such as guides. One point that is getting a lot of attention is the RIAA demand that Discord identify the accused infringers. Continue reading RIAA Alleges Popular ‘AI Hub’ on Discord Violates Copyright

Meta Platforms has unveiled Voicebox, an AI model that can produce high-quality audio clips and edit pre-recorded audio. It also uses artificial intelligence for speech generation efforts, using what Meta calls “in-context learning” to accomplish tasks it was not specifically trained for. The company says Voicebox is first in class with this type of generalized learning for audio. Untrained tasks include sampling, stylizing and editing. As an editor, it can isolate and remove sounds like car horns and background animal noise while preserving the content and style of the source audio. The multilingual model generates speech in six languages. Continue reading Meta Creates Voicebox Generative AI Model for Audio Synth

Meta Platforms has debuted what’s being called “ChatGPT for audio.” MusicGen is an AI music generator that can create tunes from natural language or song snippets. The company says MusicGen was trained on 20,000 hours of music, including 10,000 hours of “high-quality” licensed songs and 390,000 instrumental tracks. Meta released MusicGen on GitHub this past weekend, and is currently demoing the app on Facebook’s Hugging Face page. Visitors can generate tunes by describing the sound they want. Among Meta’s prompts: “80s driving pop song with heavy drums and synth pads in the background.” Continue reading Meta’s MusicGen AI Works with Language and Song Prompts

Meta Platforms continues to make progress on a mission to develop artificial intelligence that can teach itself to learn how the world works. Chief AI Scientist Yann LeCun has taken a special interest in developing the new model, called Image Joint Embedding Predictive Architecture, or I-JEPA, which learns by building an internal representation of the outside world and analyzing image abstracts instead of comparing pixels. The approach allows AI techto learn more like humans do, with their ability to figure out complex tasks and adapt to new situations. Continue reading Meta Develops Computer Vision AI That Learns Like Humans

Deezer, the global music streaming platform based in France, claims to have developed a technique for flagging — and potentially deleting — songs that use artificial intelligence to simulate the performance of popular singers. “We need to take a stand now,” Deezer CEO Jeronimo Folgueira said in an interview. “We are at a pivotal moment in music.” His company plans to “weed out illegal and fraudulent content” in an effort to protect artists. Deezer’s detection technology is still under development. It relies on AI, which Folgueira said he is not against if it is used ethically. Continue reading Deezer Says Its Tech Can Flag and Delete Deepfake AI Tunes

Meta Platforms has built and is open-sourcing ImageBind, an artificial intelligence that combines six modalities: audio, visual, text, thermal, movement and depth data. Currently a research project, it suggests a future in which AI models generate multisensory content. “ImageBind equips machines with a holistic understanding that connects objects in a photo with how they will sound, their 3D shape, how warm or cold they are, and how they move,” Meta says. In other words, ImageBind’s approach more closely approximates human thinking by training on the relationship between things rather than ingesting massive datasets so as absorb every possibility. Continue reading Meta’s Open-Source ImageBind Works Across Six Modalities

Google is readying an API and other enterprise tools for its Pathways Language Model (PaLM) — a large language model similar to GPT — to encourage developers to create chatbots and other apps using the platform. PaLM is one of Google’s most advanced systems, with the capability to generate text, images, code, video and audio from natural language prompts. Much like OpenAI’s GTP series and the LLaMA family from Meta Platforms, it is suitable for a wide variety of general tasks. To facilitate PaLM’s use for specific tasks, Google is launching the MakerSuite along with the PaLM API. Continue reading Google’s PaLM API, MakerSuite Coming to Select Developers

Chat app Discord is expanding the use of artificial intelligence on its platform, including the addition of OpenAI technology to its chatbot and moderation features. Discord says it has 150 million users across 19 million interest groups, called “servers,” that dialogue using text, audio and video chat. Discord’s Midjourney text-to-image generation group is its largest community, with in excess of 13 million members. “Harnessed properly, AI can fundamentally enhance and empower genuine human connection,” Discord CEO Jason Citron said at a press event last week, heralding “the most exciting moments in technology emerging.” Continue reading Discord Integrates OpenAI Tech, Updates AI-Driven Features

The Entertainment Technology Center@USC has released the second installment of its case study, “Fathead: Virtual Production & Beyond.” Section 2 of the four-part white paper is “Sound Mitigation: Performance Matters,” which features compelling interviews with “Fathead” co-producer Brandyn Johnson and former Sony Pictures executive Eric Rigney. The section also addresses “the challenges of recording clean dialogue on LED volumetric stages and in-camera visual effects (ICVFX) during production.” Click here to access Section 2 and the previously released Section 1, “Cloud Computing: Growth Without Bounds.” We’ll post announcements when the remaining two sections become available. Continue reading ETC Releases Next Section of Virtual Production White Paper

Spotify is adding new features that will allow for more social expression and help users discover new music, among other things. The audio streaming giant service is adding a video feed designed to recommend songs, podcasts and audiobooks via short clips, like those found on TikTok, YouTube Shorts and Instagram. “Previews,” as they’re called, allow users to swipe through content recommendations. Generated either via algorithm or configured by an artist or podcaster, the short videos are meant to encourage a deep dive into something new or saving for later.
Continue reading Spotify Launches New Video Feed to Keep Listeners Listening

Microsoft researchers have unveiled Kosmos-1, a new AI model the company says analyzes images for content, performs visual text recognition, solves visual puzzles and passes visual IQ tests. It also understands natural language instructions. The new model is what’s known as multimodal AI, which means it uses different instruction sets, from text to audio and video. Mixing media is a key step in building artificial general intelligence (AGI) that can perform tasks in a manner approximating human performance. Examples from a Kosmos-1 research paper show it can effectively analyze images, answering questions about them. Continue reading Microsoft Unveils AI Model That Comprehends Image Content

OpenAI is now allowing third-party developers integrate ChatGPT into their apps, a solution the company says will be a more cost-effective alternative. The language model can be used for more than chat, says OpenAI, which also has a new speech-to-text model called Whisper. The company is also touting gpt-3.5-turbo, calling it the “best model for many non-chat use cases.” With a major investment from Microsoft, and the eyes of the industry on it, OpenAI seems to be feeling some pressure to add earnings to the success it has as a thought leader. Continue reading OpenAI Targets Affordable AI with ChatGPT and Whisper APIs

Meta Platforms announced today that it is introducing new creative tools and features for Facebook Reels, including support for videos of up to 90 seconds, extending the previous maximum of 60 seconds. The updates arrive a few months after the company unveiled support for Instagram Reels of the same duration (news that followed TikTok’s jump in video length from three to 10 minutes in an attempt to more directly take on Google’s YouTube). Among the new creative tools include the ability to create Reels with trending templates and a “Grooves” feature that automatically syncs video to the beat of a song. Continue reading Meta Adds New Creative Tools, Features for Facebook Reels

Following several months of tests, YouTube is launching is multi-language audio track feature worldwide, with popular vlogger MrBeast helping to promote the new feature’s benefits. MrBeast, who has over 135 million global subscribers, is hoping to attract new subscribers to his channel now that the most popular videos are dubbed into 11 different languages. The multi-language audio feature allows creators to dub new and existing videos. YouTube says more than 3,500 multi-language videos have been uploaded to the site in 40-plus languages since January of this year. Continue reading YouTube Introduces Multi-Language Audio Tracks Worldwide