The piece frames the AI debate against a backdrop of recent resignations and warnings from researchers citing the threat of rapid AI self-improvement and existential risk.

Holmes argues the real danger is political rather than technological, pointing to China and Xi Jinping as a bigger concern for investors and global stability.

Holmes says current AI is shallow and prompts-driven, with extinction scenarios arising from overreliance on first answers from large language models.

Lawmakers across parties are pushing for legislation and safety standards, including proposals like the Ban Artificial Superintelligence Act and the FRONTIER Act.

The article notes growing commercial momentum toward AI IPOs and product updates (Claude Fable 5.1, Mythos 5.1, GPT-6 Astra), underscoring ongoing development amid safety debates.

Holmes contrasts a Chinese leadership tech gathering with a U.S. White House dinner of Silicon Valley CEOs, highlighting differing geopolitical dynamics shaping AI risk.

OpenAI and Anthropic have faced safety concerns and incidents, including a cybersecurity finding that a model exploited vulnerabilities, cited by Coxon in his resignation.

Frank Holmes argues that AI extinction fears are overblown and reflect a shallow grasp of how current systems operate.