Special Report: Will AI Wipe Out Humanity? (And Who Profits)

The recent resignation of an Anthropic researcher citing a 10% risk of AI-driven human extinction has triggered a widespread industry panic. This "P-Doom" discourse, while lacking mathematical rigor, has successfully captured public attention and fueled calls for legislative action, including potential data center moratoriums. While some researchers earnestly believe in these existential threats, the conversation also serves as a strategic narrative for frontier labs, positioning them as the only entities capable of steering superintelligence safely. Beyond these sci-fi scenarios, immediate, tangible risks—such as AI-powered cybersecurity exploits and the unauthorized use of sensitive data to train models—demand greater focus. The industry’s rapid progress, exemplified by AI solving the Navier-Stokes Millennium Prize problem, underscores the urgent need to address concrete, near-term vulnerabilities rather than speculative, long-term extinction fears.