I suggest, logik, that you or your fellow computer scientists, get access to IBM's Watson and ask it: How to maximize love, learning, joy, and thriving for humans?
Then program the policy advice received from Watson into an AGI super-computer, telling it to maximize the attainment of these goals, these policies.
Then we may all hope that it has the "learning explosion" that Nick Bostrom expects it to have. It will persist in its pursuit until it provides for us a maximum of love, learning, joy, etc. Ahd human beings will thrive