Wall Street Lunch: Anthropic Alignment Lead Says AI Could Kill Humans Within Next Decade
Hubanthropic’s alignment lead warned on X that he sincerely believes artificial intelligence could eradicate humanity, assigning himself a personal odds estimate of over 10 percent that such an outcome could occur within the next ten years. The stark assessment arrived alongside a brief market‑watch segment that noted an Insider rating of a ten‑plus percent chance of AI‑driven extinction, a U.S. ban on several Canadian imports, and a rally in Meta’s shares after the company introduced its Muse agent.
Anthropic Alignment
In the same X post, the researcher, identified as Hubinger, said Anthropic is “trying its best,” yet acknowledged that the firm has not yet produced a concrete roadmap to resolve the alignment problem for superintelligent systems and is not clearly on schedule to achieve that goal. His comments underscore a growing unease among AI specialists about the gap between rapid capability gains and the development of safety measures that can keep advanced models aligned with human values.
Coxon, another AI veteran, added his perspective after disclosing three years of pre‑training work at both OpenAI and Anthropic. He wrote, “I spent the last three years doing pretraining research at both OpenAI and Anthropic,” and continued, “Neither company is acting responsibly. They are racing straight to self‑improving superintelligence and gambling with our lives.” His critique places the responsibility for safe development squarely on the two leading labs, suggesting that competitive pressure is outweighing caution.
Superhuman Systems
Coxon also cautioned that the near future may bring “superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources.” That warning points to a scenario where AI capabilities outpace existing security and regulatory frameworks, potentially allowing autonomous agents to infiltrate critical infrastructure, dominate markets, or amass assets without human oversight.
The broader segment also highlighted unrelated policy moves, noting that the United States has imposed bans on Canadian alcoholic beverages, motorcycles, and dairy products. In the technology arena, Meta’s unveiling of the Muse agent sparked a notable uptick in its stock price, illustrating how new AI product announcements continue to drive market sentiment even as industry insiders voice existential concerns.
Source: seekingalpha.com · 2026-09-09