A “clean” image and video generation model, trained exclusively on “ethically sourced” data, is being positioned for enterprise-grade productions in Hollywood next year as an alternative to offerings from tech giants accused of data scraping. The new model, dubbed “Marey,” after early cinema pioneer Étienne-Jules Marey, will be available in January or February 2025.
Source: ‘Clean’ AI Video Model to Launch in Early 2025 Targeting Hollywood Clients (EXCLUSIVE)
Like many organizations in the music industry, Merlin supports “AI products that aid human creativity, or provide new opportunities for artists to create and collaborate in developing new original works,” wrote the organization in a mission statement shared with Billboard. But it strongly opposes “any product, regardless of its purpose, that has been trained on Merlin members’ music without permission.”
Music AI, an AI-powered audio technology company, has partnered with Audible Magic to help film and TV companies manage music licensing. Through Music AI’s cinematic stem separation technology, users will be able to break down video soundtracks into individual components like dialogue, background noise, and music. They will then be able to swap in new music when needed.
Source: Music AI partners with Audible Magic to create music-swapping tool for TV, video and film
The rosiest outlook on AI deployment envisions an invigoration of the production landscape by bringing down costs in areas related to conceptualization, visual effects and animation, just to name a few. The trade-off? Chunks of the Hollywood workforce gone, mostly those in postproduction who are staring down the barrel of potential automation.
The Writers Guild of America‘s east and west chapters have sent a strongly-worded letter to the heads of the major Hollywood studios, criticizing them for inaction as artificial intelligence appears to be taking copy-written scripts and using it to “plagiarize stolen works.” The WGA officers go on to criticize the studios for doing “nothing to stop this theft. They have allowed tech companies to plunder entire libraries without permission or compensation.”
As major media companies eye consolidation, Ferguson has signaled that he will be more lenient that Lina Khan, the FTC chair under President Joe Biden. Punchbowl News last week reported on a pitch made to Trump’s team for Ferguson, and it included the argument that he would stop Khan’s “war on mergers.” He also said he would “protect freedom of speech and fight wokeness” and that he would “fight back against the trans agenda.”
Source: Donald Trump Appoints Andrew Ferguson As Next Chair Of Federal Trade Commission
Of note, Murray said that audiobooks continue to outsell e-books in terms of total revenues and are helping to offset declines from lagging e-book sales. In terms of AI, the CEO declined to discuss specifics of licensing deals, like the one it signed with Microsoft for AI model training, citing NDAs, but stressed that it’s early days for AI licensing, calling it a “fascinating time” in the industry.
Source: HarperCollins CEO touts Spotify’s audiobooks entry, AI’s impact on publishing
Paul McCartney has backed calls for laws to stop mass copyright theft by companies building generative artificial intelligence, warning AI “could just take over.” The former Beatle said it would be “a very sad thing indeed” if young composers and writers could not protect their intellectual property from the rise of algorithmic models, which so far have learned by digesting mountains of copyrighted material.
Source: Paul McCartney warns AI ‘could take over’ as UK debates copyright laws
Apollo Research tested six frontier models for “in-context scheming” — a model’s ability to take action they haven’t been given directly and then lie about it. After being told to achieve a certain goal “at all costs,” evaluators introduce contradictory information to see if the model will remove any restrictions in its path and deceive developers.
Source: OpenAI’s o1 lies more than any major AI model. Why that matters
OpenAI launched its video-generating tool, Sora, on Monday. But the company’s opting not to release a key feature for most users pending further testing. The feature in question generates a video using an uploaded photo or footage of a real person as a reference. OpenAI says that it’ll give a “subset” of Sora users access to it, but that it won’t roll out the capability broadly until it has a chance to fine-tune its “approach to safety.”
Source: Why OpenAI is only letting some Sora users create videos of real people