Cyborgs are desirable because they solve the "control problem" of AGI. Essentially, if you create a superintelligent AI, how do you make sure it doesn't outwit

Cyborgs are desirable because they solve the "control problem" of AGI. Essentially, if you create a superintelligent AI, how do you make sure it doesn't outwit you and all other humans in order to do terrible things? There aren't any plausible methods that I'm aware of to accomplish that goal. So you sidestep it by not making any AI that's more intelligent than humans. Unfortunately there also aren't good ways to make sure that no one makes AI that's more intelligent than humans (plus arguably superintelligent AI could be an extremely good thing if it doesn't turn into Skynet). So the remaining obvious option is increasing human intelligence at the same time as we improve AI -- or even better, merge humans and AI together.