Sure we do, "we" just refuse to acknowledge this, keeping a 100 year old debate alive for no good reason.
Macroscopic objects follow microscopic rules. That's that. There's no further debate. There can be none.
I can't even begin to describe how absurd it is to argue anything else. It's like... saying with a straight face that software doesn't "really" follow the rules of boolean algebra if it has enough lines of code. That somehow once a program gets "big enough", it can transcend truth tables and somehow go analog or something.
It's like a mathematician saying that really big equations, the kind that span several pages stop following the rules of algebra.
Get it? It's just... insane. The rules of the universe are the rules for all things in it. They apply to everything, at all scales, at all times.
If they don't, they're not the rules!
QM and GR are approximations. The real rules are unknown to us. It just happens that "in the small", QM seems to be a good approximation that is consistent with experimental results. And "in the large" the same holds for GR. Neither of them works for everything though.
That just means we need to find a better approximation. That's not insane.
Bits are not an approximation of software "in the small". They are the real building block. We know that because we made them. First made them, then observed how they behave. But QM is a theory created by first observing. Physics is a natural science trying to understand what we observe. Math and CS are not. The objects they care about are conceived by us and observed second.
This is nowhere near as certain as you make it out to be. Take Conway's Life. With 4 simple "rules of the universe", you can create a series which to the best of our knowledge cannot be predicted from those rules. Clearly this series is a "thing" within the simulated universe, and clearly it emerges from the universal rules, but the rules don't give us any insight about it! The only known way to "predict" it is to let it run: in other words, it's irreducibly complex.
Causality can look different at different scales. It's well established that simple low-level rules can generate extreme if not irreducible complexity. This doesn't mean the bottom-level rules don't apply everywhere; it just means their descriptive/predictive utility is not necessarily preserved when you zoom out. Is a prime-finding algorithm best described by the mechanism of the computational substrate? Any number of substrates could suffice. We can best predict its output by thinking at a higher level of abstraction.
You can absolutely reversibly compute conways game of life state, and you can compute it forward as well (after all, it is a game). That's prediction.
> It's well established that simple low-level rules can generate [...] irreducible complexity.
Simple rules can create a very complex emergent system, but that doesn't mean it cannot still be reduced to it's component rules. That's what makes them rules and not just guidelines.
Notice I said irreducible, not irreversible. Yes, you can reverse the computation; but there's no known way to "shortcut" the forward process to predict the outcome of the series I mentioned any faster than simply letting it run. Letting it run is not prediction in the sense I mean here. By "predict" I mean foretell in advance what the system will do without needing to let it run.
> Simple rules can create a very complex emergent system, but that doesn't mean it cannot still be reduced to it's component rules. That's what makes them rules and not just guidelines.
I agree, with a minor caveat about language: you can describe a system which may display complex emergent behavior in terms of its underlying rules, but doing so is not guaranteed to give you useful information about (i.e., allow you to "predict", in the sense described above) the emergent behaviors. In contexts like these, to "reduce" typically means to describe in terms of a lower-level formalism while preserving predictive ability, AFAIK. In other words, the low-level formalism provides full information about the behavior of the entire system, such that you don't need to observe the system to know what it will do. In the Life example this is not the case.
Incidentally, Greg Egan wrote 2 short stories on basically that premise: Luminous and Dark Integers.
Furthermore, leaving behind direct observations which could perhaps be waved away with discussions of probabilities, we have one big problem: there is no gravity in QM and we have no idea how to account for it or curved spacetime in QM.
So for now, we have one working model for the macroscopic world (general relativity) and one for the microscopic world (QM), but the two are mathematically incompatible, they can't be simultaneously true, and we have not yet found an experiment which contradicts either of them.
> there is no gravity in QM
The weakness(es) of any one particular theory doesn't in an way disprove that macroscopic objects follow the same rules that microscopic objects do.
Just because MySQL is bad doesn't mean that the relational model is false, or that databases are pointless.
Just because the current theories of GR and QM aren't easily extended to all regimes doesn't meant that there is some sort of hard boundary where the rules change. Our theories of the world don't affect how it works. The boundaries of our theories are not boundaries of the world.
You can't fall of the edge of the world because the maps only go so far...
Yes, absolutely agreed. My point was simply that we don't know how QM extends to macroscopic objects, not that there must be some hard boundary (though we can't exclude the possibility that there exists some hard boundary at some level of energy, just as we know that the Standard Model doesn't describe matter at certain high energies).
Until we have some unification of GR and QM, we can't say for sure that QM describes the macroscopic world, just as we can't say that GR describes the behaviors of particles. Most likely we will at some point find such a model, and find out exactly how QM applies to large systems - perhaps with some limits to uncertainty, similarly to how c acts as a limit to speeds.
Sure it does - gravitational lensing, etc.
As a programmer, I wholly endorse this notion.
It literally does sometimes, just ask anyone that's programmed software designed to be resistant to bit flips caused by cosmic radiation.
Well, for now we don't really know how the rules of QM apply to macroscopic objects. The exact physical interpretation of wave-function collapse (if any) is still a matter of speculation, with all options still being on the table - maybe there is no collapse (e.g. many worlds theory), maybe there exists a collapse when interacting with large enough systems (measurement - the Copenhagen interpretation), maybe collapse is a physical process that happens at precise scales (Roger Penrose seems to believe something like this), maybe the wave function is a physical wave (pilot wave theory) and there is no collapse in another way.