In something of a first, Google has launched a new AI application for books that publishers seem to actually like, thanks to its apparent respect for copyright protections. Today, the company introduced Expert Intelligence (EI), a feature that allows someone reading an e-book purchased on Google Play’s e-bookstore to interact with the title—by asking questions about the book, taking a quiz after reading it, or producing an infographic of key data—through Google’s AI research tool, Gemini Notebook.
Source: Google Launches AI Tool to Add Interactivity to E-books
Sony Music Publishing and Warner Chappell Music have joined forces to sue Anthropic, home of Claude, over what the publishers call “one of the largest and most blatant ongoing thefts of intellectual property in history.” The complaint also names Anthropic Co-Founder and Chief Executive Officer Dario Amodei and Co-Founder Benjamin Mann as individual defendants. It was filed on Friday (August 28) in the US District Court for the Northern District of California.
Imagine a scenario where a documentary filmmaker, in the course of making the documentary, captures some damning footage of corporate malfeasance, which she wishes to share with an investigative reporting organization anonymously. Should we be concerned that mandates on AI watermarking might reveal who she is, even if she’s not using AI at all? This doesn’t mean that watermarking shouldn’t be used, but rather, as WITNESS notes, we should be aware of the risks, and seek to counter them.
SoundCloud announced that it will begin rolling out direct music purchases, now in beta, on the streaming platform direct from artist pages. Starting Wednesday (Aug. 26), around 200 Artist Pro creators will be able to sell songs directly from their profiles on SoundCloud, providing an additional source of revenue for artists. “Artists should be able to build their audience in one place, on one platform,” Eliah Seton, SoundCloud’s CEO, said in a statement. “SoundCloud has always been where artists find their first fans and build real communities.
Streaming platform Twitch and its parent company Amazon are facing a proposed class action lawsuit filed by a Connecticut-based streamer over claims that the companies harvested creators’ content without informing or compensating them to train Amazon’s artificial intelligence programs. According to the filing, streamer and class representative Warren Pandiscia alleges that Amazon began scraping video streams for training purposes before Twitch updated its terms of service to address AI data use, which included an opt-out option for creators.
The full array of patents extends beyond music to audio, video, film, animation, images, and text, as well as applications involving name and likeness, according to MIH. Digital services can license them through an online platform that MIH said is designed for widespread adoption. Meanwhile, UMG, its affiliates, artists and songwriters – plus other participating rightsholders – will be able to use the portfolio to protect their works against unlicensed and unauthorized generative AI uses.
In an interview with the New York Times, Dr. Dre and Interscope Records co-founder Jimmy Iovine spoke about artificial intelligence’s recent foray into music production, with the producer characterizing those who oppose AI song generation tools as “afraid of learning new things.” “I don’t see it as a threat. I think the only people that see it as a threat are the people who have trouble creating,” Dr. Dre said.
AI companies are now required to make model outputs detectable as AI-generated under Article 50 of the EU AI Act – or, in other words, watermarked. Responses, however, have been mixed. While Anthropic stated it has seen no notable drop in usage, sentiment has reportedly been shifting, as per Forbes. Some users are concerned that the watermark can appear on text they have only used Claude to edit, making it indistinguishable from wholly generated text and resulting in reputational decline.
In the study, published in Nature, the MIT researchers Zheng Dai and David Gifford set out to to test whether a generated image can be traced back to a single piece of training data. They were looking specifically at diffusion models, the systems most often used for generating images and video. Their finding? It comes down to how big the training data set is. They found that the more data a model is trained on, the harder it becomes to attribute its output to any particular piece of training data.
Australia passed legislation on Thursday that will force tech giants to pay millions of dollars in levies if they fail to strike commercial deals with local media outlets for news on their platforms. The News Bargaining Incentive taxes the companies 2.5% on their advertising revenues unless they strike agreements. Proceeds from the scheme would be directed to local Australian news outlets, whose content helps drive user engagement and advertising revenue on the tech firms’ platforms.