You say, building ontologies with a specific intent will actually lead to intelligence at a similar level of current AI?

This would be a severe blow to the bitter lesson, and generally vindicate the AI researchers of the 80s that put a bet on symbolic processing