Some fun questions to consider

[BradKML]

1. Can this model of reinforcement learning handle adversarial teaching (e.g. drafting trick questions)?
2. Is the student training method compatible with Absolute Zero (self-play/self-study) and Intuitor (coherence checking)?
3. How can RLT be expanded to open-world learning (e.g. browsing internet and reading books/papers for truth determination)?