Neural networks already have their own internal knowledge representations. They just aren’t capable of learning new knowledge (without expensive re-training or fine-tuning).
Inference is cheap, training is expensive. It’s a really difficult problem, but one that will probably need to be solved to approach true intelligence.
In the way that they're trained to complete tasks from users, can they be trained to complete tasks that require usage of a memory storage and retrieval mechanism?
Where does fine-tuning sit in this? How easily are existing models able to be fine-tuned for new use cases, such as specifically legal or medical texts?
Comments
Neural networks already have their own internal knowledge representations. They just aren’t capable of learning new knowledge (without expensive re-training or fine-tuning).
Inference is cheap, training is expensive. It’s a really difficult problem, but one that will probably need to be solved to approach true intelligence.
In the way that they're trained to complete tasks from users, can they be trained to complete tasks that require usage of a memory storage and retrieval mechanism?
Where does fine-tuning sit in this? How easily are existing models able to be fine-tuned for new use cases, such as specifically legal or medical texts?