From "Artificial General Intelligence" to "Applied General Intelligence" (parallel channel: t.me/agichat ) -- how to make it safe, and make sure it advances the humanity's ability to collectively define and pursue goals.
This channel shall be a bit of a mix of "Artificial General Intelligence" and "Applied General Intelligence". As described in t.me/agichat, we had invented a way to pose the problem of universal optimization thousands of years ago by inventing an equation model. We (our society) already is a kind of general intelligence. However, it is not yet capable to solve some of the problems, such as the causes of people's deaths, and so on. Just it being "general" doesn't necessarily imply omnipotence. I think, it just means that, just like a "Universal Turning Machine", it can be general in some domain, given enough resources...
Likely, the capabilities of AGI will be tied directly with the resources allocated for goals. Worth to point out, that perhaps any AGI will organize and prioritize ideas around goals.