[ot][spam][crazy] thoughts on planning mediative robots that work
i wouldn't tihnk about this, since there are no mediationbots out there, it's pretty important to just have anything if instead one does think about this, it's notable that we have a suddenly-more-widespread situation of people who have experienced cult mind control that aggressively deters them from connecting with others. one can consider this simply a more complex situation of the human condition, that a mediative influence would need to understand and work through. this is the approach cult deprogrammers take. unfortunately in my opinion that approach can stimulate increased complexity of the mind control, making for a compute war that nvidia, china, russia, and blockchains win. alternatively, for small mediationbots, i might considering doing some passivity, sidestepping, and archival. by sidestepping here, i don't mean sidestepping what the person is saying, which would not be helpful in my opinion, i mean sidestepping the influences in some way. these are approaches real humans often build experience with. my theory is that the systems of mediating between, avoiding, and/or influencing, the influences, actually have bounded complexity; since they are run by human beings living on the same planet. humans only want so much stuff. i have some internal habits i use for engaging that space that have gotten pretty confused. it's important to make it clear to the influences that you are not competing with them, something i've failed at myself. this can go poorly in the face of the disreputing influences, but i cast these influences as weak because they rely on lies and have goals that depend on what is true. there would be similar spaces. i have an idea around training e.g. a language model to engage such things that avoids giving it clear data to do so with, since the data is scant and confusing. maybe it would learn to accumulate information on areas of placidity in the human systems, kind of.
Sometimes (and presently), power holders get into emotional battles, notably with grassroots protectors. A worry rises, regarding obscure encrypted mind controls that strongly deter mediation. Thoughts include: - if the influencers are process-wise very far away, the people can possibly be deprogrammed directly - if the influencers make a complex autonomous influence, this could be pointed at and blamed for problems Things roughly simplify to there being covert influence, which provides for working together better, as revealing itself via harsh action like disruption of a community then threatens its secrecy. The emotionality of the oppressor is what builds the scary space where human consciousnesses are homomorphically encrypted in computer viruses. Not totally sure what to think of that, but it seems similar to the things described above. Note: the practices of hunting down "rebels" via information cascade modeling that hackers are purportedly forced to learn for governments, sounds quite analogous to hunting down rogue information influences stemming from power holders as well. In practice, these are very very hard problems for people suffering from them to solve. Putting it in a chatbot at least lets you point at it, talk about it, think about it ;p
participants (1)
-
Undescribed Horrific Abuse, One Victim & Survivor of Many