Possible Future types of AI include:
1) AI that possesses consciousness, sentience, and self-awareness, similar to humans.
2) AI that can understand human emotions, beliefs, and intentions, and use this understanding to interact with humans more effectively. It is not yet a reality.
3) AI that possesses human-level intelligence and can perform any intellectual task that a human can. If it took over AI research from humans, it could theoretically develop a ----
4) Superintelligent AI
"What is it about future AI systems that might make them capable of causing catastrophic harm? One milestone that is commonly cited as a threshold at which large-scale risks become serious is Artificial General Intelligence, or AGI. By AGI, we mean AI that can perform most or all economically relevant tasks as effectively as any human."
Narayanan, Arvind; Kapoor, Sayash. AI Snake Oil: What Artificial Intelligence Can Do, What It Can’t, and How to Tell the Difference (p. 150). Princeton University Press. Kindle Edition.