> possibly that our brains allocate more resources to learning than to inference It's likely in brain inference is learning. If you want a technical analog it'

> possibly that our brains allocate more resources to learning than to inference It's likely in brain inference is learning. If you want a technical analog it's like a conversation in LLMs. Previous tokens do affect the currently generated. I.e. it's inference time learning, well known and widely used.