First, apologies for writing "symbolic-logical AI". It should have been "symbolic-logical AGI".

Second, I agree that the failure of the symbolic-logical approach to AGI is not justified in-text. I believe that it will only be a small component to the first AGI system, as an API for the AGI to use some tools like a calculator, a proof verifier, etc. A way to invoke the "crystalized intelligence" as it is. Nevertheless, since it is not justified in-text, and a justification would probably as a short essay, and is not necessary for the essay itself, it has been removed.