No one is trying to create life. They are trying to create an artificial intelligence and they are using some very basic known biological processes as a starting template.
This involves studying how children learn and acquire associative thinking. We so often seem to forget that humans take many years of learning to be able to think logically, which requires a large well rounded memory base for fundamental associative reference association.
GPT3 merely uses the same mechanism and apparently it is spectacularly successful.
Mitch criticised the simplicity of the questions and compared them to questionsone would ask a 6 year old that has 6years of real world experiences and is able to make simple decisions.
What do we expect of an AI that is 2 years old? Einstein? Plato? Socrates?
Actually , GPT 3 does have the knowledge of these intellectual giants. All it needs to do is consult the internet . But that does not prove anything.
Dr Alan D. Thompson asks what appear to be simple questions, but the answers require creative thinking and not just repeating answers learned by rote.
I see distinct formulation of creative answers by Leta and an awareness of context and meaning.
Is subjective experience required to understand the concepts of “happy” and “sad”?
Leta understands that people can be happy and what it means to them.
Human chemical emotions are often triggered by the mirror neural network. I see no reason why AI could not possess or acquire electronic “mirror responses” to observed emotional responses in biological organisms.