Abstract:
According to Business Insider, former Anthropic researcher Jacob Coxon detailed his differences with the AI giant during a Q&A event on the X platform on Tuesday. Coxon previously became popular for his resignation warning, believing that AI may destroy humanity in the future.
Coxen
Coxson said Anthropic leadership is wrong to use a "consequentialist philosophy" to explain why it must participate in pushing the frontiers of AI research.
"If this contest is unavoidable, you should not be participating in it," Coxon wrote.
Coxson specifically singled out Anthropic for its push for recursive self-improvement (RSI), the process by which AI models can improve themselves and build their own successor models.
"They instigated the recent race around RSI to a large extent because they believed that this race was inevitable. It should be noted that OpenAI had to abandon a lot of baggage such as Sora because Anthropic was going full force at the time." Coxon said.
Coxson elaborates on his differences with Anthropic
Coxson also said that Anthropic’s leadership is too wary of the Chinese and U.S. governments. “They do not believe that negotiations are possible.”
Anthropic CEO Dario Amodei has repeatedly expressed unease: If Anthropic is the only cutting-edge lab to slow down development, other AI companies may adopt different security measures.
"If we develop too slowly, I still think that the wrong people will end up controlling this technology," Amodei told CNN recently. "This will also greatly increase the possibility of things going wrong."
Comments