Image by Areps.com, from Unsplash
Hunger Strikes Target AI Labs Amid Rising Worries About “Superhuman” Intelligence
- Written by Kiara Fabbri Former Tech News Writer
- Fact-Checked by Sarah Frazier Former Content Manager
Two activists have started hunger strikes in front of Anthropic and DeepMind offices, demanding the companies to halt their development of advanced AI systems.
In a rush? Here are the quick facts:
- Reichstadter demands Anthropic halt frontier AI development immediately.
- He is surviving only on water, electrolytes, and multivitamins.
- Reichstadter previously staged a 15-day hunger strike over climate issues.
Guido Reichstadter spent seven days without food outside Anthropic’s San Francisco office building as he continued his protest. The activist told Business Insider he will maintain his hunger strike until Anthropic responds to his requests.
“DeepMind, Anthropic and other AI companies are racing to create ever more powerful AI systems. Experts are warning us that this race to ever more powerful artificial general intelligence puts our lives and well being at risk, as well as the lives and well being of our loved ones,” he wrote on X .
Indeed, recent research warned that AI firms are racing toward human-level systems without proper safety plans. A new Future of Life Institute report gave no company a grade higher than a C+ in existential risk preparedness.
Experts fear artificial general intelligence could emerge within a decade, yet safety strategies remain “unacceptable,” leaving society vulnerable as AI models become harder to understand.
Reichstadter, founder of Stop AI, argues that pushing toward “superhuman” AI poses existential risks. “In the concrete world in which we’re living right now, all of the frontier labs are racing as quickly as they can to fully general superhuman systems. That’s what needs to stop,” he said to Business Insider.
On his first day of protest, he handed a letter to Anthropic CEO Dario Amodei requesting him to stop frontier model development.
Amodei himself has warned about AI’s economic impact, saying in May: “AI may eliminate 50% of entry-level white-collar jobs within the next five years.”
He noted to Business Insider that he previously conducted a 15-day hunger strike in Miami over the climate crisis that did not harm his health. He was also arrested for chaining shut the doors of OpenAI’s San Francisco office and is set to face trial this month.
His protest has inspired Michael Trazzi, a 29-year-old former AI safety researcher, who has been fasting outside DeepMind’s London headquarters for three days. Business Insider reports that Trazzi said his demand is simple: “My only ask is, concretely, I want Demis to say that he would not release any more frontier models if the other frontier AI labs were to also stop doing so.”
Photo by Patrick Collins on Unsplash
Warner Bros. Discovery Sues Midjourney For Using Copyrighted Characters
- Written by Andrea Miliani Former Tech News Expert
- Fact-Checked by Sarah Frazier Former Content Manager
Warner Bros. Discovery sued the company Midjourney on Thursday, claiming that the AI startup has been using copyrighted characters in its image-generation system, including its “most popular and valuable fictional characters” such as Batman, Superman, Wonder Woman, Scooby-Doo, and Bugs Bunny.
In a rush? Here are the quick facts:
- Warner Bros. Discovery sued the company Midjourney on Thursday over the use of its popular copyrighted characters.
- The lawsuit presents evidence showing how Midjourney’s AI system generates images and videos of its characters.
- Batman, Superman, Wonder Woman, Scooby-Doo, Bugs Bunny, Tweety, The Powerpuff Girls, and Rick and Morty are among the characters mentioned in the complaint.
According to The Verge , Warner Bros. Discovery included several examples of how Midjourney has been reproducing “unauthorized derivatives,” downloadable videos and images, and distributing them to users without permission.
“Only Warner Bros. Discovery has the right under U.S. Copyright law to build a business around reproducing, preparing derivative works, distributing, publicly displaying, and performing images and videos featuring its copyrighted characters,” states the lawsuit. “Midjourney thinks it is above the law.”
Throughout the formal complaint, Warner Bros. Discovery presents evidence showing how Midjourney’s AI system generates its characters—such as Scooby Doo, Batman, Wonder Woman, Tweety, Superman, Batman, Rick and Morty, and The Powerpuff Girls—through prompts in different scenarios, even when the user has not explicitly requested the copyrighted character.
“Without any consent or authorization by Warner Bros. Discovery, Midjourney brazenly dispenses Warner Bros. Discovery’s intellectual property as if it were its own,” states the complaint. “Midjourney could easily stop its theft and exploitation of Warner Bros. Discovery’s intellectual property.”
Warner Bros. Discovery is suing for damages and requested the court to stop Midjourney from generating, copying, and distributing its copyrighted content.
Midjourney—which recently partnered with Meta on video and image generation technology—has also been sued by Disney and Universal under similar claims earlier this year.
Artists and creative companies have long complained about the use of their work to train AI models. However, this year, multiple judges have sided with tech companies in copyright cases. In June, a San Francisco judge ruled in favor of Anthropic in a case over the use of purchased and pirated copyrighted books, while another judge ruled in favor of Meta in a case brought by 13 authors who claimed the company used their books to train its AI model.
