AI Experts Warn: We’re Closer Than Ever to Existential Risk
Experts say AI could destroy humanity by decade's end. New warnings spark global debate on safety and control.
POLICY WIRE — City, Country — A former Anthropic researcher has sparked a global conversation after claiming that artificial intelligence is already capable of causing catastrophic harm, even if it requires human intervention.
Jacob Coxon, who previously worked at OpenAI, recently left his position at Anthropic and shared his concerns in a widely viewed post that gained over 100 million views. He argued that the industry is racing toward self-improving superintelligence, which could pose an existential threat to humanity within the next decade.
While some experts believe true superintelligence remains distant, others warn that current AI systems are already powerful enough to cause significant damage. Vu Tran, a former Meta researcher, claimed that leading companies could cripple entire nations with just a few strategic actions. His comments have added urgency to the growing debate over AI safety and regulation.
📄 POLICY WIRE WHITEPAPER PUBLISHED: PAKISTAN’S NATIONAL SECURITY POLICY PRIORITIES
Industry leaders like Dario Amodei of Anthropic and Sam Altman of OpenAI have called for slower development of AI models to prevent potential disasters. However, international cooperation remains a challenge, as countries like China continue to invest heavily in their own AI advancements.
Some researchers have proposed technical solutions, such as implementing a “kill switch” to shut down rogue AI systems. In the UK, lawmakers have introduced legislation aimed at creating a safety net for critical infrastructure, though the proposal was recently rejected by the government.
Reporting by Policy-Wire (PW)
