If with “AGI” you mean “the ability of an intelligent agent to understand or learn any intellectual task that a human being can” (https://en.wikipedia.org/wiki/

If with “AGI” you mean “the ability of an intelligent agent to understand or learn any intellectual task that a human being can” (https://en.wikipedia.org/wiki/Artificial_general_intelligence https://en.wikipedia.org/wiki/Artificial_general_intelligenc...), not something (vastly) surpassing humans, and add “limited processing ability”, I don’t think AGI would be able to crack something we cannot crack with the 8-ish billions humans we have. Also, many of the large problems we have (wars, climate change) aren’t intellectual but social, to start with.