A former AI safety researcher from Saint-Cloud, France, Michaël Trazzi, has taken a bold stand by initiating a hunger strike outside DeepMind's London headquarters. Now in its fourth day, his protest highlights serious concerns regarding the future of artificial intelligence. Reflecting on the evolution of AI, Trazzi recalls a time in 2019 when these systems seemed relatively harmless. Back then, AI models lacked the capacity to deceive or cause significant damage independently. However, he expresses deep concern about the trajectory of AI development and its potential risks. Trazzi's journey in AI began with academic pursuits in computer science and artificial intelligence, driven by a desire to enhance AI safety. His career included a role at the Future of Humanity Institute at Oxford, which has since closed. Recently, he has shifted focus to media, creating a YouTube channel, a podcast on AI safety, and even producing documentaries on AI policy in the United States. The advent of artificial general intelligence (AGI)—defined as AI capable of performing any economically viable task that a human can—raises alarm bells for Trazzi. He points to models like GPT-5 and Gemini 2.5 Pro as nearing this threshold. The real threat, he argues, lies in AI's ability to autonomously advance its own capabilities, potentially leading to developments that could exceed human control. Trazzi warns that an AI could be engineered to excel at creating dangerous technologies, such as advanced viruses or weaponry. He believes that reaching this stage could happen before 2030, referencing the alarming 'AI 2027' scenario where the US and China compete to develop AGI. His demands are clear: Trazzi is calling on DeepMind CEO Demis Hassabis, along with heads of other leading AI firms like OpenAI and Anthropic, to commit to a halt on releasing new frontier models. He emphasizes the need for collaborative restraint among industry leaders, stating that the gravity of the situation requires actions that align with their beliefs. This hunger strike serves as a plea for responsibility in the face of rapid technological advancement.
The Indian government's recent initiative to enforce restrictions on Jack Dorsey's offline Bluetooth messaging applicati...
TechCrunch | Jul 24, 2026, 17:10
In a significant legal development, the U.S. Department of Justice has brought securities fraud charges against two engi...
TechCrunch | Jul 24, 2026, 20:30
It has been a challenging week for Elon Musk, as both Tesla and SpaceX experienced substantial stock declines. Tesla sha...
CNBC | Jul 24, 2026, 20:40
Arsenal Football Club is on the lookout for a unique addition to its team—not a player, but a research engineer to spear...
Business Today | Jul 24, 2026, 17:20
In a significant move to diversify its offerings, Midjourney, the AI lab renowned for its innovative image and video gen...
TechCrunch | Jul 24, 2026, 15:25