A recent analysis of AI Overviews found that they were accurate approximately nine out of 10 times. But with Google processing more than five trillion searches a year, this means that it provides tens of millions of erroneous answers every hour (or hundreds of thousands of inaccuracies every minute), according to an analysis done by an A.I. start-up called Oumi.
Major music publishers and the NMPA are asking a Texas federal court to throw out X’s antitrust lawsuit, calling it a baseless attempt at retaliation. The music companies argue that X’s conspiracy theory rests on a single word in an email, while adding that their massive DMCA takedown campaign was not a sham but fully protected by the First Amendment.
Source: Music Publishers Ask Court to Dismiss X’s ‘Weaponized DMCA’ Antitrust Suit
And, as with previous introductions of new technology into Hollywood, from digital film to email, AI is percolating from the bottom up, starting with the assistant class — on track to become industry standard as today’s underlings (those that survive the continual layoffs, that is) rise to positions of power. Confronted with larger workloads and a shrinking headcount, AI — both the kind officially approved by companies and more surreptitious uses — has made its way via support staff into essential Hollywood workflows.
Source: Hollywood Assistants Are Using AI Despite Their Better Judgment
The guild reached a tentative agreement on Saturday with the Alliance of Motion Picture and Television Producers. The deal will run for four years — instead of the typical three-year term — a significant benefit to the studios, which will get an extra year without the possibility of a strike. Details were not immediately forthcoming, but the deal was expected to include a major cash infusion into the guild’s teetering health fund, which has bled $200 million over the last four years.
Source: WGA Reaches Surprise Deal With Studios a Month Before Contract Expires
The statement, published on Thursday (April 2), came one day after WMG announced it had entered into a definitive agreement to acquire the independent music platform, which specializes in digital music distribution, rights management, royalty accounting, and real-time analytics. IMPALA acknowledged that the deal reflects well on the independent sector’s capacity for innovation, but argued that it continues a pattern in which independent infrastructure is absorbed by major label groups
Source: IMPALA says WMG’s Revelator acquisition is ‘bad news for artists, fans, and Europe’s diversity’
Last week’s Supreme Court decision in Cox Communications reshaped the piracy liability landscape, creating new urgency for site-blocking legislation in Congress. This could be addressed by Senator Thom Tillis and Representative Zoe Lofgren, who have been working on bicameral legislation that would require ISPs and DNS resolvers to block foreign pirate sites under court order, TorrentFreak has learned.
Source: U.S. Lawmakers Work on Unified Site-Blocking Bill to Counter Online Piracy
On the facts of the case, the Chinese court, in a case of first impression, held that copyright did not subsist in six sets of prompts submitted to Midjourney and hence the copying of those prompts could not infringe copyright. However, the court suggested that there could be copyright in a more elaborate and specific set of prompts that meets copyright’s standard for originality.
Source: Copyright in AI Prompts: Chinese Court Ruling on Generative AI and Originality
Voice AI company ElevenLabs has quietly released an iOS app called ElevenMusic that can be used to create music with AI and discover AI-generated music. The new app, which was listed on the App Store for a few weeks and finally released on April 1, suggests ElevenLabs wants to be more than just a voice model company. The company sees tools that use AI to create music and other mediums as a way to grow and protect itself from the eventual commoditization of AI audio models.
Source: ElevenLabs releases a new AI-powered music-generation app
The executive order’s instructed development of “best practice guidance for departments and agencies to appropriately watermark” all “AI-generated or significantly manipulated images or video.” Longer term, making AI watermarks the norm across the board would, of course, be a sweeping step – and a step with multiple considerations beneath the surface. How would doing so affect the growing collection of creations incorporating, but not solely resulting from, generative AI?
Source: CA’s Newsom Issues AI Executive Order After Trump Push Against ‘Cumbersome State AI Laws’
These gig AI trainers – who upload everything from scenes around them to photos, videos and audio of themselves – are at the frontlines of a new global data gold rush. As Silicon Valley’s hunger for high-quality, human-grade data outpaces what can be scraped from the open internet, a thriving industry of data marketplaces has emerged to bridge the gap. From Cape Town to Chicago, thousands of people are now micro-licensing their biometric identities and intimate data to train the next generation of AI.
Source: Thousands of people are selling their identities to train AI – but at what cost?