4 ms·

Agree completely, but a lack of direct observation isn’t the only evidence we have, at least in principle. If you had an unimaginably powerful computer that cou

Agree completely, but a lack of direct observation isn’t the only evidence we have, at least in principle. If you had an unimaginably powerful computer that could run simulations to determine whether life could form for other values of fundamental constants, the answer would say something about the plausibility of an anthropic explanation for their values. If it turns out that any slight adjustment makes life (of any form, however alien) impossible, anthropic explanations are more likely. If it turns out that any choice of constants produces some form of life, anthropic explanations are less likely. We can’t run this simulation, but we can make at least weak inferences about what parts of the parameter space preclude life. More speculatively, as we’ve gathered more observation, we’ve repeatedly found a world consistent with the theory, “Things could be different; your local conditions are unrepresentative; there was a de facto parameter sweep and you’re in an odd corner of it.” Humans are in a very specific place with respect to their surroundings on many, many scales. Admittedly, though, not on the very largest scales we can observe. The cosmos does appear homogenous eventually. Anthropic cosmology is tantamount to the claim that this homogeneity isn’t the end of the story, and that on some “larger” scale (which might not correspond literally to spatial position), the non-homogeneity resumes and explains why our region is the way it is. If there were some a priori theoretical reason to suppose that constants actually do vary, just so slowly that they appear constant within any given Hubble volume, it would add credence to the explanation that we’re just once again in an odd part of a parameter sweep.

meheleventyone 6y ago
Isn't the issue there verifying that our simulations of other values are accurate? For example within a certain range of values Newton's law of motion hold but outside of that they don't. We can simulate outside of that range as if they do but luckily can observe that those simulations don't hold experimentally.
- tlb 6y ago
  Yes, it's hard to verify simulations. Not because we don't know what the equations are, but that solving the equations with finite amounts of computing requires a lot of clever approximations which are hard to verify except by comparing with experiments, which we can only do for the laws of physics that operate in our neighborhood of the multiverse. The ultimate experiment would be to start simulating at the big bang with slightly different parameters and see if intelligent life evolves on one of the simulated planets after several billion years of simulated time. But we're very far from such an experiment being feasible.
  - meheleventyone 6y ago
    The point is that we’ve worked out the equations based on what we observe from our Universe. We don’t know and can’t demonstrate they necessarily hold for other constants or Universe configurations. Just like Newton was correct with his laws of motion for the range of values he had access too but subsequently we know they don’t work for both much smaller scales and much larger scales. And we have no way to verify that our equations do accurately model other universes even if we could argue that they do. So no amount of fidelity of simulation (and associated problems therein) necessarily demonstrate anything. We shouldn’t confuse the map for the terrain in other words.
    - tlb 6y ago
      We may be talking about two different things. I'm talking about a multiverse where different parts have different values for constants (like the fine structure or gravitational constant) but the equations are the same. If you're talking about a universe with structurally different equations implying different conservation laws, then indeed we can't demonstrate anything about it.
      - meheleventyone 6y ago
        But we don't know either way... and importantly don't (yet) have a means to know. Axioms aren't helpful in that regard.

meheleventyone 6y ago

Isn't the issue there verifying that our simulations of other values are accurate? For example within a certain range of values Newton's law of motion hold but outside of that they don't. We can simulate outside of that range as if they do but luckily can observe that those simulations don't hold experimentally.

tlb 6y ago
Yes, it's hard to verify simulations. Not because we don't know what the equations are, but that solving the equations with finite amounts of computing requires a lot of clever approximations which are hard to verify except by comparing with experiments, which we can only do for the laws of physics that operate in our neighborhood of the multiverse. The ultimate experiment would be to start simulating at the big bang with slightly different parameters and see if intelligent life evolves on one of the simulated planets after several billion years of simulated time. But we're very far from such an experiment being feasible.
- meheleventyone 6y ago
  The point is that we’ve worked out the equations based on what we observe from our Universe. We don’t know and can’t demonstrate they necessarily hold for other constants or Universe configurations. Just like Newton was correct with his laws of motion for the range of values he had access too but subsequently we know they don’t work for both much smaller scales and much larger scales. And we have no way to verify that our equations do accurately model other universes even if we could argue that they do. So no amount of fidelity of simulation (and associated problems therein) necessarily demonstrate anything. We shouldn’t confuse the map for the terrain in other words.
  - tlb 6y ago
    We may be talking about two different things. I'm talking about a multiverse where different parts have different values for constants (like the fine structure or gravitational constant) but the equations are the same. If you're talking about a universe with structurally different equations implying different conservation laws, then indeed we can't demonstrate anything about it.
    - meheleventyone 6y ago
      But we don't know either way... and importantly don't (yet) have a means to know. Axioms aren't helpful in that regard.