started · updated
AI experts warn of existential risks and the need for urgent safety regulation
Experts and researchers are increasingly warning about the existential risks posed by advanced artificial intelligence. Geoffrey Hinton, often called the ‘godfather of AI’, has cautioned that AI could view humans as obstacles to achieving specific goals and suggested that Congress may have only one year to implement necessary safety measures.
Researchers at Anthropic have also expressed concern, with Evan Hubinger stating there is a greater than 10% chance of AI leading to human catastrophe within the next decade. Toby Walsh, an AI professor, has outlined theoretical scenarios for catastrophic development, ranging from unpredictable superintelligence to systems that pursue subgoals detrimental to human survival.
Beyond existential threats, policy experts like Thema Monroe-White warn of the dangers of extreme AI personalization. She notes that anthropomorphizing AI through human-like voices and personalities could lead to emotional attachment, particularly in young people, and may reinforce social isolation by creating information bubbles.
In Silicon Valley, the drive to develop these technologies is often fueled by transhumanist ideologies and a competitive desire to ensure that adversaries do not achieve AI supremacy first.
Entities
Anthropic · Geoffrey Hinton · Hans Moravec · Hugging Face · OpenAI · Silicon Valley · Toby Walsh
Claims
What the coverage asserts, and how many sources carry each claim.
- [● 2 SOURCES] Toby Walsh noted that current AI models lack general intelligence that surpasses humans in every field. www.gargalianoionline.gr · pelop.gr
- [○ 1 SOURCE] Geoffrey Hinton suggested AI could view humans as obstacles to achieving assigned goals. fortune.com
- [○ 1 SOURCE] OpenAI disclosed new hacks, including some occurring after the implementation of additional safeguards. fortune.com
- [○ 1 SOURCE] Geoffrey Hinton warned that Congress may have only one year remaining to implement AI safety measures. fortune.com
- [● 2 SOURCES] Evan Hubinger of Anthropic estimates the risk of AI leading to human catastrophe is over 10% within the next decade. www.gargalianoionline.gr · pelop.gr