What is currently called the AGI field will probably bear no fruit, perhaps except for the end-game when it borrows then-sufficiently powerful tools from more productive areas of research (and destroys the world). “Narrow AI” develops the tools that could eventually allow the construction of random-preference AGI.
What is currently called the AGI field will probably bear no fruit, perhaps except for the end-game when it borrows then-sufficiently powerful tools from more productive areas of research (and destroys the world). “Narrow AI” develops the tools that could eventually allow the construction of random-preference AGI.