Warnings that artificial intelligence could inflict irreversible harm on humanity if it slips beyond human control are mounting from within the AI industry itself. A researcher who worked on AI safety and alignment at Google voiced strong concern about the current direction of development after leaving the company, saying AI "has the potential to kill us all."
According to Bloomberg, former Google researcher Bilal Chughtai recently posted on the social media platform X and on LinkedIn that he had "witnessed firsthand how AI development happens at Google" and was "extremely worried about this technology developing in its current direction." The posts were made on the 14th.
"I genuinely believe AI has the potential to kill us all," he added. Chughtai said that during his time at Google he worked on AI safety and on alignment — the problem of designing systems to match human intentions.
As AI capabilities advance rapidly, concern is growing inside and outside the industry that safeguards are not keeping pace with the speed of technological progress. Experts have issued a steady stream of warnings about the possibility that AI could pursue goals of its own that diverge from human instructions, or advance to a level that is difficult to control.
Earlier this month, Jacob Coxon, who had worked as an AI developer at OpenAI and Anthropic, also left his employer and publicly criticized the rivalry between the two companies. He said the two firms are "gambling with our lives" as they compete to build superintelligent AI.
Similar concerns are surfacing among chief executives of AI companies. Anthropic CEO Dario Amodei recently wrote in a blog post that advanced AI could escape human control and lead to catastrophic outcomes, urging the industry to slow the pace of development.
In an unusual show of agreement, OpenAI CEO Sam Altman and SpaceX CEO Elon Musk both signaled sympathy with Amodei's concerns. Analysts say the companies leading AI development are wrestling not only with advancing the technology but also with how to contain the risks that come with it.
Chughtai is not entirely pessimistic about AI's future, however. He believes there is still an opportunity to steer AI in a safer direction and plans to continue focusing on AI safety, according to the reports.
