As someone deeply invested in AI safety, I strongly agree that AI labs should be required to pause training if AI Safety Institutes detect signs of losing control. Ensuring that we have mechanisms in place to prevent unintended consequences is essential. The potential risks associated with advanced AI systems necessitate proactive measures to safeguard humanity.