I really want this to be true, but honestly it's really hard. What makes you think this won't be eaten too within the next year based on the current s-curve-if-not-exponential we are on?
My guess is that (as in most fields) the advancements will be more convoluted and surprising than the simple idea of "we now have AGI".