Emerging forms of artificial intelligence, including “world models”, self-improving AI, agentic AI, and the possibility of artificial general intelligence (AGI), may significantly extend current capabilities by simulating aspects of real-world environments or matching or surpassing human capabilities. These systems could deliver substantial economic and societal benefits, such as scientific discovery and advanced automation, but also increase risks including systemic misuse, loss of human oversight, the possibility of models being uncontrollable, and challenges in ensuring alignment with human values. A proposed POSTnote would examine areas of leading AI research, what AI systems may be in use in the future, their capabilities and applications, and currently proposed safeguards.