In a striking analogy, Jacob Coxon, a former researcher at Anthropic, has likened the development of superintelligent AI to the emergence of an alien species whose intelligence is beyond human comprehension. Speaking in an interview, Coxon expressed that this notion is indeed unsettling, highlighting the potential dangers of creating an intelligence that surpasses human capabilities. He noted that, even without a specific scenario in mind, the prospect of superior intelligence raises significant concerns.
Coxon recently exited the AI industry, citing the rapid pace at which companies are moving toward developing self-improving systems. He warned that this rush could lead to a loss of control over such technologies. In light of these warnings, Dario Amodei, the head of Anthropic, has suggested a slowdown in the development of the most powerful AI models to ensure that safety measures can keep pace with technological advancements. This call for caution has garnered public support from prominent figures, including OpenAI's CEO Sam Altman and entrepreneur Elon Musk.
Coxon's fears are echoed by Evan Hubinger, the head of Alignment Science at Anthropic, who previously estimated that the likelihood of superintelligent AI leading to human extinction within the next decade exceeds 10%. This figure, he stressed, represents his personal assessment.
Meanwhile, OpenAI has postponed its initial public offering amid ongoing safety concerns surrounding AI. In a related development, SpaceX is set to launch AI systems based on NVIDIA's Vera Rubin NVL72 into orbit by 2027. This optimized version is expected to serve as the foundation for the first generation of Starmind satellites, designed for performing AI computations in space.
These developments signify a critical moment for the AI market and may prompt competitors to enhance their safety protocols and rethink their developmental strategies in light of growing public and expert concerns.
Informational material. 18+.
