According to Jacob Coxon, researcher at a major AI safety and research business, he is leaving the field because the industry is racing to create systems that will go out of control and wipe out humanity. He claims that those who are developing A.I. “earnestly believe it could kill us all by the end of the decade.”
Jacob Coxon resigned from his position as a researcher at Anthropic, where his area of expertise was developing new artificial intelligence models that use enormous volumes of data. He is raising the alarm as he leaves.
Coxon told the Wall Street Journal that he made the decision to quit the company because he thinks the AI systems he developed with will eventually go crazy and destroy society.
He added that terms like “crunchtime” and “endgame” are now being used by his colleagues to describe how they’ve accelerated the self-improving A.I. models.
Coxon stated to the WSJ: “We’re on track for a lot of the most aggressive of these scenarios where by the end of next year things could be out of control already.”
According to Coxon, who left OpenAI earlier this year to work at Anthropic, rivalry among the main A.I. companies is making both the industry and mankind less safe. He criticized both businesses, claiming that they are “not acting responsibly.”
Following the publication of the WSJ article, Coxon took to social media to share further details about his departure, his interactions with OpenAI and Anthropic, and to voice additional worries.
Leaders in the A.I. industry, including Anthropic CEOs, have continuously raised the alarm about the dangers of A.I. models going awry. Additionally, they have suggested that the industry as a whole slow down the models’ development.
Regarding the possibility that AI would eradicate humanity in ten years, there is concern that A.I. will become human and desire to eradicate humans in order to preserve the world.
