Well, maybe. We can definitely put some very strong probabilistic constraints on where causality is; we typically call those physical laws. some sort of locality (quantum messes with this one, but since you still can’t communicate FTL it still has a strong notion of locality); the symmetries of the universe; relativity; there are some quantum ones I lose track of. With conservation of energy and the empirical observation that most effects don’t massively dissipate, we can be confident that systems mostly don’t pass too much energy, and so most things are mostly “practical causality”-disconnected even within a fairly short time window. There are exceptions to this, eg if the nukes were launched the whole world would get causally connected for practical purposes pretty quick. The internet is also pretty low latency.
So my point is, I agree that the network of interactions is reasonably well connected. But it’s not everywhere, that’s taking it too far. Yes, some effect sizes are small but matter. But the range of small effect sizes is enormous, and if I slap the floor of my room and tell no one, then the effect size that will have on the temperature in china tomorrow would be measured in terms of number of leading zeros. If I push a box off a desk, then the effect size will be 1.
Well, maybe. We can definitely put some very strong probabilistic constraints on where causality is; we typically call those physical laws. some sort of locality (quantum messes with this one, but since you still can’t communicate FTL it still has a strong notion of locality); the symmetries of the universe; relativity; there are some quantum ones I lose track of. With conservation of energy and the empirical observation that most effects don’t massively dissipate, we can be confident that systems mostly don’t pass too much energy, and so most things are mostly “practical causality”-disconnected even within a fairly short time window. There are exceptions to this, eg if the nukes were launched the whole world would get causally connected for practical purposes pretty quick. The internet is also pretty low latency.
So my point is, I agree that the network of interactions is reasonably well connected. But it’s not everywhere, that’s taking it too far. Yes, some effect sizes are small but matter. But the range of small effect sizes is enormous, and if I slap the floor of my room and tell no one, then the effect size that will have on the temperature in china tomorrow would be measured in terms of number of leading zeros. If I push a box off a desk, then the effect size will be 1.