Dario Amodei, the chief executive of Anthropic, recently issued a warning regarding the rapid evolution of artificial intelligence. In an open letter, he stated that within the next 6 to 12 months, AI-driven botnet swarms could potentially gain control over the entire internet. Amodei estimates that such persistent botnets could cause hundreds of billions of dollars in economic damage.
The executive emphasized that the scale of potential harm is likely to grow if AI systems continue to advance without the implementation of necessary safety guardrails. He noted that current models are already capable of assisting with cyber operations, weapons engineering, and military intelligence, which could lower the barriers for adversaries lacking human resources.
The perspective shared by Amodei is echoed by former Anthropic AI scientist Evan Hubinger. Hubinger recently stated on social media that he believes there is a greater than 10% chance of human extinction within the next decade due to AI. He expressed concern that there is currently no established plan to solve the alignment problem for superintelligence.
Despite these warnings, the industry remains divided on the path forward. Some experts argue that more advanced AI technology could also provide better safeguards and automated detection of malicious activity. Furthermore, there is a concern that halting development might be counterproductive if less responsible actors continue to advance their own systems.
