Vasco Grilo comments on AI Safety Newsletter #3: AI policy proposals and a new challenger approaches