Rapid artificial intelligence innovation has pitted creatives and technologists in opposition, and in the last week alone, nearly every day has brought new skirmishes.
On Monday, actress Scarlett Johansson threw OpenAI chief Sam Altman in hot water over his company’s use of a chatbot voice that sounded similar to hers, after she declined to record her own voice for the AI. “I was shocked, angered and in disbelief that Mr. Altman would pursue a voice that sounded so eerily similar to mine that my closest friends and news outlets could not tell the difference,” she said in a statement. OpenAI denied any intentional resemblance and paused use of the chatbot’s voice.
On Wednesday, OpenAI cut a deal with News Corp to train its large language models on the publisher’s news properties, including the Wall Street Journal, the New York Post, The Sun, and a dozen more. The Information’s editor-in-chief Jessica Lessin called the move a “fatal error” because it works against publishers’ own business interests, as AI companies build products that could replace news sites. A New York Times lawsuit filed in December against OpenAI for using its content to train ChatGPT is ongoing.
On Thursday, Bloomberg reported that Google and Meta are talking with major Hollywood studios about licensing content to train their AI models. Warner Bros. Discovery is said to be open to sharing some of its titles, while Netflix and Disney are unwilling to license their content.
The friction isn’t new. Consent and compensation around the use of AI were huge negotiating points driving last year’s strikes by Hollywood actors and writers. But as tech companies race to create more advanced models—especially in the difficult and finicky world of video generation—the conflict has only heated up. AI companies require mass amounts of data to train their technologies, and creative industries are rich sources of original material. Voice acting, writing, film, photography, music—every art form appears to be a target for AI. And many creatives have no interest in their work being used as training data.
As my colleague Sharon Goldman covered in last week’s Data Sheet, more than 200 musicians signed an open letter demanding protections against AI, calling it an “assault on human creativity.” The artists include Katy Perry, Nicki Minaj, and Billie Eilish, who have some cultural sway on the matter.
Lesser-known artists have taken to Reddit and X, formerly known as Twitter, to share recommendations on how to protect their work. One self-made guide from an artist on X shows Instagram and Facebook users how to prevent their profiles from being used to train Meta’s AI models. Posted on Thursday, it has been shared more than 5,800 times and saved by more than 8,000 accounts on X.
All signs point to a continued push from AI companies to acquire more data, indicating more rifts between tech companies and artists to come. No creator is safe.
We’ll be discussing this topic and many more at Fortune’s annual, invite-only Brainstorm Tech summit in Park City, Utah, from July 15-17. Joining us for a conversation about the future of AI in Big Tech are Google’s chief scientist Jeff Dean and Amazon’s head scientist Rohit Prasad. Sequoia managing partner Roelof Botha will discuss big bets in venture capital, Runway cofounder and CEO Cristobal Valenzuela will talk about how AI is changing filmmaking, and Agility Robotics CEO Peggy Johnson will address how human-centric robots are revolutionizing industries. Want to join us? Register here.
With that, here’s the latest tech news.
Rachyl Jones
Want to send thoughts or suggestions to Data Sheet? Drop a line here.
The rest of today’s Data Sheet was curated by David Meyer.
Data Sheet will be back on Tuesday, following the Memorial Day holiday in the U.S. on Monday.
NEWSWORTHY
First FCC gen AI fine. And the Federal Communications Commission’s first fine for misuse of generative AI goes to: Steve Kramer, a “political consultant” who deepfaked President Joe Biden’s voice in robocalls to New Hampshire voters. The Associated Press reports that Kramer’s fine is $6 million, while Lingo Telecom, which allegedly transmitted the calls, received a $2 million fine.
Nvidia’s China struggle. Nvidia has reportedly had to cut the prices of its most advanced AI chip in China (which U.S. sanctions have forced to be less powerful than the AI chips it can sell elsewhere) due to weak demand. According to Reuters, Nvidia now has to sell its H20 chip at a price set to undercut Huawei’s Ascend 910B, which is the most powerful AI chip made by a local player.
xAI funding. Silicon Valley heavyweight VCs Lightspeed, Andreessen Horowitz, Sequoia, and Tribe have all joined a “close to $6 billion” funding round for Elon Musk’s AI startup xAI, the Financial Times reports. Musk hasn’t quite reached his target, though. In related news, OpenAI—surely Musk’s leading target with his xAI venture—has released most of its past employees from a non-disparagement agreement that it had forced on them by tying it to their equity and their exit contracts.
Spotify Car Thing outrage. Spotify will intentionally deactivate its Car Thing gadgets in December. It launched and soon stopped making the device—a smart music player for the car—a couple years ago, but some people bought them and are none too pleased that Spotify will brick them. Ars Technica reports that some Car Thing owners have been pleading with the company to open-source the device, so hobbyists can make and install their own firmware rather than having to throw the thing away.
IN OUR FEED
“Samsung does not seem interested in enabling repair at scale.”
—iFixit CEO Kyle Wiens is done with Samsung, The Verge reports. The repair-parts service and the smartphone giant are discontinuing a pioneering two-year partnership that made it easier for customers to access spare parts of their handsets. Wiens reportedly doesn’t like Samsung’s parts pricing, nor its habit of forcing people to buy a new screen if they just want to replace the battery.
IN CASE YOU MISSED IT
Nvidia’s landmark performance hasn’t been seen ‘in the history of capitalism,’ says tech CEO, by Marco Quiroz-Gutierrez
Google’s AI search wants you to glue cheese to your pizza. It’s just the tip of its bad-idea iceberg, by Chris Morris
The White House is pleading with Big Tech to shut off the firehose of sexually abusive AI deepfakes—‘If you’re a teenage girl, if you’re a gay kid, these are problems that people are experiencing right now’, by the Associated Press
OpenAI’s week of chaos has reopened a festering wound at the $80 billion startup that was supposed to have healed, by Sharon Goldman
Amazon readies Alexa for its second act—but it has an uphill battle, by Sage Lazzaro
SEC completes abrupt U-turn to approve Ethereum ETFs in ‘politically driven’ move, by Niamh Rowe
BEFORE YOU GO
Very good doge. Rest in peace, Kabosu, the Shiba Inu dog whose image became a meme, the inspiration for the Dogecoin cryptocurrency, and a $4 million NFT. The BBC reports that Kabosu died today at the age of maybe around 18—her owners, who adopted her from a puppy mill, aren’t quite sure when her real birthday was.
