RL combined with Transformers and some form of symbolic logic may get us closer to AGI but it is an inefficient approach compared to going for neuromorphic AI to get closer to the human level brain.

→ View original post on X — @deeplearn007