Trump defends AI approach, criticizes Anthropic CEO and 'conspiracy theorists'

In recent days, the internet has been abuzz with concerns over the potential AI apocalypse. Jacob Coxon, a former employee at OpenAI and Anthropic, sparked the conversation by quitting the industry, stating that neither company was acting responsibly. Many in the AI community echoed his concerns, admitting that discussions about the potential extinction of humanity are common behind closed doors.

In response, OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei—both former colleagues of Coxon—joined the discourse. Despite their past rivalry, the two tech leaders agree on the need to slow down AI development to prevent a global catastrophe. Altman addressed the issue in two posts on social media. In the first, he explained that OpenAI was now treating development and evaluation with the same level of care, including slowing down the overall pace of development.

He emphasized that "pacing" was different from "stopping," as safety cases and monitoring come with significant costs. However, he warned that the pace should not be slower than necessary, as recklessness and unchecked capabilities could be more dangerous. In a follow-up post, Altman went further, expressing his fears of losing control of the future to AI.

He stressed that he and his team were firmly on the side of humanity, ensuring that AI served people. He also highlighted the danger of an AI monopoly, where an extraordinarily powerful AI could be used by a single person or company to impose their worldview on the world, leading to a dystopian outcome. To avoid these scenarios, Altman concluded that AI companies and governments must find a "narrow middle path" where no single entity controls the technology.

Amodei echoed similar sentiments in an essay on his website titled "We Must Pace the Frontier." He advocated for slowing down AI development without halting it entirely. Amodei warned that building AI too quickly could be reckless, while not building it could deprive humanity of benefits or hand the technology to authoritarian powers.

He proposed a three-step plan to "pace the frontier," including embedding evaluators to ensure accountability, democratic coordination between AI companies in democratic countries to create common safety standards, and global coordination through the U.S. government and other democratic nations to impose safety standards on authoritarian regimes.

Amodei acknowledged the challenges of implementing these measures but emphasized the importance of trying to ensure safety while still achieving AI's benefits and addressing geopolitical dilemmas.