Suddenly, this thread turns into a physics meeting.
What I meant by "knowing the state of the entire universe" is that if you know in what state the entire universe was one nanosecond (or any other length of time) ago, that still isn't enough to causally infere in what state that specific particle is now (one nanosecond later). What I'm doing here is basically acknowledging quantum randomness and the absence of perfect causality.
The thing with Shrodinger's cat is that even if you know the decay rate of whatever radioactive stuff you are using, what happens in practice is random. If you leave the cat for one minute inside the box, you know what are the odds it is dead, but you don't know whether it is or not, and you could very well design the experiment in a way that the only way to know is to observe inside the box. When you think about it, everything in life is like that because everything is linked, strongly or not, to quantum physics and randomness.
For one, why would only a human eye count as an observer, and the cat itself doesn't count? Also with the standard interpretation we get a "Schrödinger's cat inside a box, inside a box, inside a box" scenario if, for example, there's someone else outside the room when you open the box. From their perspective, you found either a dead cat or a live cat, and until they walk in the room and see which one is true, and the wavefunction collapses. And from the perspective of someone in another room... and outside the building.. and outside the city... what's outside all of those boxes that's causing the wavefunction to collapse?
The uncertainty is an individual thing, it's not the universe witnessing you specifically are trying to fuck with him and suddenly says "this particle will collapse its wavefunction !" when you observe it. The experiment can even be seen from the perspective of the device that detects radioactive decay. When it detects a gamma ray or whatever, its uncertainty about when the phenomenon actually happens dissappears, but yours does not.