i only need to show a toddler what an apple is once, and he will recognize a apple in every color just fine, i dont need to train him with tons of data and erro

i only need to show a toddler what an apple is once, and he will recognize a apple in every color just fine, i dont need to train him with tons of data and error correction, and tell him "use this relu function here and sigmoid here". A toddler can also start questioning why some apples are red and other greens? and he can come with new ideas of his own, can LLM do that? no, because their answers depends on the data they have been trained for. Im not saying it's not the future, it is, we just need the hardware, once there's a hardware cheap enough that replicates what an human does with his hands, then the days working at a factory will be over. The current AI's are perfect for copycating behavior.