TL;DR warning: Stable, low-enenergy, high entropy states are thermodynamically preferred and complex emergent systems like life and sentience are more efficient at producing those. This is why the naive calculations of fine-tuning grossly underestimate the number of possible universe in which something with enough wits to make those calculations will emerge.
I don't love the multiverse hypothesis, largely because it is a hypothesis and currently has no evidence. I accept that it's possible, but it's inellegant. There's a simpler answer that requires fewer assumptions and can be done with a single observable universe. Complex emergent behaviors are favoured by thermodynamics and if you get enough of those for long enough then what you have is sentience. It may be wildly alien sentience that we'll never be able to recognize, but that's what you get for mucking about with fundamental constants.
Step 1: Nature tends toward stability. This is just the Second Law of Thermodynamics and it can be derived from nothing more complicated than basic logic and statistics. There's only one way to be a blender full of nitroglycerine but an enormously large number of ways to be a slowly cooling cloud of blender fragments and gas. Give them equal space on a dart board and you'll never even see the spot with the nitroglycerine, much less hit it with a random shot. Probably just as well.
Step 2: Negative feedback loops are typically long-lived. Positive feedback loops are flashy and a great way to liberate a lot of energy, but they grow so uncontrollably that they inevitably run up against another rule that kills them off. Supernovae, avalanches, and nitroglycerine in a blender are all exciting but brief. They also don't get you anywhere. In all of those cases the immediate products have almost exactly the same energy as the original starting material. All you've done is exchange nuclear/gravitational/chemical potential for heat/kinetic energy. By contrast, when two or more rules inhibit one another you get negative feedback loops that tend to self-regulate at steady states. Red dwarf stars may be unexciting but they'll be shining trillions of years after the supernova has faded. Unsurprisingly, they're the majority of stars in the galaxy. Glaciers can exist for millions of years and span continents, making avalanches look laughable by comparison. And most people would prefer a blender full of milkshake for obvious and sane reasons.
Step 3: Complex emergent behaviors are more efficient at creating entropy. Emergent behaviors are a bit weird, but you can think of them fairly simply as a set of simple rules which, when repeated again and again, begin exhibiting behaviors that aren't explicitly stated by the original rules. It isn't that these systems start breaking their original rules, it's that they start acting in new ways on top of those. Yes, that's strange, but we observe them everywhere that we've looked from sociology straight down to pure mathematics. In this universe they're nested several layers deep: fundamental laws interact to give rise to physics, the laws of physics gives rise to chemistry, the rules of chemistry gives rise to life, the way that life behaves gives rise to sentience, and the shit that sentience gets up to gives rise to societies. Emergent behaviors are better at turning free energy into entropy simply because there are so many more possible ways for something to be.
As an analogy, take three very different roller coasters. In all three cases the cars are winched up to the top of the ride and then released.
So, given all of that, nature prefers complex emergent behaviors, not just because they're interesting, but because they're thermodynamically favourable systems for arriving at stable, low-energy states more efficiently. Fine-tuning arguments don't take that into account, focusing on anthropocentric carbon-based and human life, and failing to account for all the other possible ways that you could arrive at something sentient if conditions were different. Honestly, I'm hoping that we do live in a multiverse and that we find a way to explore it, because then we're going to see some really wild shit!
I don't love the multiverse hypothesis, largely because it is a hypothesis and currently has no evidence. I accept that it's possible, but it's inellegant. There's a simpler answer that requires fewer assumptions and can be done with a single observable universe. Complex emergent behaviors are favoured by thermodynamics and if you get enough of those for long enough then what you have is sentience. It may be wildly alien sentience that we'll never be able to recognize, but that's what you get for mucking about with fundamental constants.
Step 1: Nature tends toward stability. This is just the Second Law of Thermodynamics and it can be derived from nothing more complicated than basic logic and statistics. There's only one way to be a blender full of nitroglycerine but an enormously large number of ways to be a slowly cooling cloud of blender fragments and gas. Give them equal space on a dart board and you'll never even see the spot with the nitroglycerine, much less hit it with a random shot. Probably just as well.
Step 2: Negative feedback loops are typically long-lived. Positive feedback loops are flashy and a great way to liberate a lot of energy, but they grow so uncontrollably that they inevitably run up against another rule that kills them off. Supernovae, avalanches, and nitroglycerine in a blender are all exciting but brief. They also don't get you anywhere. In all of those cases the immediate products have almost exactly the same energy as the original starting material. All you've done is exchange nuclear/gravitational/chemical potential for heat/kinetic energy. By contrast, when two or more rules inhibit one another you get negative feedback loops that tend to self-regulate at steady states. Red dwarf stars may be unexciting but they'll be shining trillions of years after the supernova has faded. Unsurprisingly, they're the majority of stars in the galaxy. Glaciers can exist for millions of years and span continents, making avalanches look laughable by comparison. And most people would prefer a blender full of milkshake for obvious and sane reasons.
Step 3: Complex emergent behaviors are more efficient at creating entropy. Emergent behaviors are a bit weird, but you can think of them fairly simply as a set of simple rules which, when repeated again and again, begin exhibiting behaviors that aren't explicitly stated by the original rules. It isn't that these systems start breaking their original rules, it's that they start acting in new ways on top of those. Yes, that's strange, but we observe them everywhere that we've looked from sociology straight down to pure mathematics. In this universe they're nested several layers deep: fundamental laws interact to give rise to physics, the laws of physics gives rise to chemistry, the rules of chemistry gives rise to life, the way that life behaves gives rise to sentience, and the shit that sentience gets up to gives rise to societies. Emergent behaviors are better at turning free energy into entropy simply because there are so many more possible ways for something to be.
As an analogy, take three very different roller coasters. In all three cases the cars are winched up to the top of the ride and then released.
- Ride #1 is the simplest possible design, one car, no track. On release it plummets straight down. At the instant of impact the car has simply traded gravitational potential for kinetic energy and little energy has been lost to entropy. Any entropy produced will result from hitting the ground hard and is an entirely separate process.
- Ride #2 is only slightly more complex. It's a long, straight track 2000 m long with barely enough slope to keep moving. This one has a nice balance between gravity and friction but in the end it's nearly as simple as ride #1. Gravitational potential is converted to kinetic energy and then to heat in the wheels and rails. It's reasonably well-distributed through space and time, but it's still all one type of energy.
- Ride #3 is the one that you know and love from childhood. It goes up and down, around high-gee turns, and occasionally inverts. Unlike the previous two rides it'll scatter output energy in three directions and will liberate heat, motion, screaming, and the occasionally unfortunate up-chuck. This is the more complex ride with lots of different states that the car can be in, so it's no surprise that it doesn't go as fast as Ride #1 or as far as Ride #2 but manages to turn a lot more free energy into entropy.
So, given all of that, nature prefers complex emergent behaviors, not just because they're interesting, but because they're thermodynamically favourable systems for arriving at stable, low-energy states more efficiently. Fine-tuning arguments don't take that into account, focusing on anthropocentric carbon-based and human life, and failing to account for all the other possible ways that you could arrive at something sentient if conditions were different. Honestly, I'm hoping that we do live in a multiverse and that we find a way to explore it, because then we're going to see some really wild shit!
