The plan to use AI to solve its own safety risks has a critical failure mode: an unlucky ordering of capabilities. If AI becomes a savant at accelerating its own R&D long before it becomes useful for complex tasks like alignment research or policy design, we could be locked into a rapid, uncontrollable takeoff.

Top AI labs like Anthropic publicly state that slowing down AI development would benefit society. However, they are caught in a strategic trap: a unilateral pause is unviable. Without a global agreement, any lab that pauses simply allows less cautious competitors to seize the lead, potentially making the ecosystem less safe.

A pause on training new, more capable AI models could paradoxically increase risk. It would halt progress at the few, relatively safety-conscious frontier labs, allowing less scrupulous competitors to catch up. Meanwhile, compute stockpiling would continue, making any subsequent capability leap even faster and more dangerous.

In the high-stakes race for AGI, nations and companies view safety protocols as a hindrance. Slowing down for safety could mean losing the race to a competitor like China, reframing caution as a luxury rather than a necessity in this competitive landscape.

AI leaders aren't ignoring risks because they're malicious, but because they are trapped in a high-stakes competitive race. This "code red" environment incentivizes patching safety issues case-by-case rather than fundamentally re-architecting AI systems to be safe by construction.

Leaders at top AI labs publicly state that the pace of AI development is reckless. However, they feel unable to slow down due to a classic game theory dilemma: if one lab pauses for safety, others will race ahead, leaving the cautious player behind.

CEOs from leading AI labs like Google DeepMind and Anthropic have publicly stated they would prefer to slow down development to address safety concerns. However, they feel compelled to continue the race because if they pause unilaterally, less cautious competitors, including state actors like China, will not.

A fundamental tension within OpenAI's board was the catch-22 of safety. While some advocated for slowing down, others argued that being too cautious would allow a less scrupulous competitor to achieve AGI first, creating an even greater safety risk for humanity. This paradox fueled internal conflict and justified a rapid development pace.

Aza Raskin reveals the internal strategy of leading AI labs is not to avoid danger, but to race towards it. Their plan is to reach the 'cliff'—the point where AI becomes uncontrollably powerful—as fast as possible, seize the resulting 'weapon,' and use it to stop all competitors.