Belief Reasoning with Subjective Logic
folk.uio.no
folk.uio.no
This statement contradicts itself:
> We can assume that an objective reality exists, as "das Ding an sich" (the-thing-in-itself) in the pholisophy of Kant, but our perception of that reality will always be subjective.
in that the latter clause is an objective statement about perception necessarily, by its assumption, derived through said limited subjective means.
"Do not rely on your thinking in order to know the world"
The subjective/objective distinction is problematic. Obviously, knowing agents are always subjects -- something has to be doing the knowing -- but knowledge of that presupposes that the subject is able to know the world as it is in the first place.
Suppose you look through a bucket of polluted water. It's difficult to see through it. Heck, it's difficult to tell what's even IN the bucket because the water is murky. If the water is clean, and you shine a light into it, then, things in it are visible.
The analogy is to the degree of truthfulness of an individual's consciousness. The higher the truthfulness, the higher the degree of precision with which they can make confirmations about what exists as it is.
> However, in the long term it's detrimental to the stability of global civilisation when we hold and promote different and conflicting subjective truths.
If it's referring to "subjective truths" as identity beliefs, then yes, it could be an accurate statement. But if "subjective truths" is referring to more malleable beliefs decoupled from identity, then I'd say it's overstating things. Conflict that can be resolved with a positive outcome for all involved is better than no conflict, as a lack of conflict can lead to stagnation.
I don't think the author means that _every_ belief held by any person must ultimately become "objective" (and, therefore, aligned with that of all the rest of humanity). For instance, I don't see how it is bad for human civilisation if I think I look ravishing in red, whereas everyone else thinks I look like a pillock.
Luckily we dont need to know whether something is objectively true for it to be useful. And useful is really all you need.
The reason many people get scared when you answer these questions with "no" is that it would mean the complete decoupling of politics from reality. Debates about policy have a long history, and they're not harmful. But when one side starts also making up their own facts, it becomes impossible to reason, and a Venezuela-style decline might follow.
Logical reasoning with transitive trust networks could obviously be useful to re-establish a common reality. But while it might convince those that answered the two questions above with 'no' that the statistics are really compiled in good faith, no amount of reasoning will help with those answering "yes" to the first, and "no" to the second.
You might, for example, trust your spouse. You have no problem giving them your car keys, because you believe they are unlikely to steal the car.
Now a stranger asks you to borrow your car. You would probably not give them the keys.
But then you learn that the stranger is a good friend of your spouse. Because you trust your spouse, and they trust their friend, your trust in the stranger just grew, and you would probably consider giving them the car.
Using many such chains, with some discounting as the length increases, you could, for example, create a trust score for a news story.
Fundamentally, this would seem to do little more than work to entrench already established forces who would tend to the score the highest, or equivalently have the resources to the game the system most effectively.