To handle a scenario I'm working with, I took a shot at explaining some of the difficulties in merging incompatible egos: An incomplete (beta, delta, etc.) fork of a person is an imprecise copy of their brain structures. Either the whole structure is coarsened and simplified, or specific parts are extracted out to construct the fork. To produce something which can function as an intelligent being, such forks are outfitted with "psychological prostheses" in software. A specialized delta fork, for example, doesn't need the richly textured emotional life of its originator, but it needs some context for such things or it cannot function. One factor which interferes with merging incompatible egos is the necessity of rationalization. Human beings act in certain ways, dictated by past behavior, personality, circumstances and so forth. However, they must also see their behaviors as the rational outcome of a decision-making process. An ego which attempts to merge with a fork whose behaviors are considerably at odds with its own standards is forced into a situation where it cannot rationalize actions it now believes itself to have taken. A potential solution is to excise the portions of the brain that would interfere with such a merge. This arguably results in a new ego being formed, rather than any essence of the previous ego/egos involved being carried through the merge process, and risks insanity. If attempted on two incompatible alpha forks, irreversible madness is the most likely outcome. Someone intimately familiar with the psychology of two distinct egos might conceivably assemble an ego from their components over time. In the case of a fully-formed ego attempting to integrate an inferior fork, the ego's "real" brain will replace the psychological prostheses which supported the fork's functionality. However, there is necessarily some cost to the ego's own brain structure. If the TITANs or other hyperintelligent AIs have ever attempted such experiments, they certainly didn't publish their data. At present, there isn't enough economic incentive to justify the overwhelming risks of this sort of experimentation (as far as anyone knows).