Yet note that this was not always so - for much of human history, the progress of science walked on a thin line to avoid falling into heresy, on one side, while trying to retain its disdain for dogmas and preconceived notions of a creator, and of a supernatural sphere that could not be empirically accessed. But in the XXIth century, it is religion that fights the uphill struggle, in an attempt at protecting the idea of creation from successful scientific attacks to its foundations, through explanation of those mysteries that once constituted the attractiveness of supernatural "explanations".
For that reason it may appear odd if I take a walk on that wild side here. Of all, me - an atheist, a reductionist scientist who believes nothing until he can double-check it with a Monte Carlo simulation. But you know what the nice thing of walks on the wild side is: they allow you to remind yourself why you chose the other side long ago.
The thought of this Sunday morning, as I took a (real) walk and breathed fresh air among daisies and quacking ducks, was on whether our Universe came out so successfully, marvelously complex, and hosting highly-desirable intelligent life (me) by pure chance, or by spontaneous emergence of complexity through some yet-to-be-understood principium, or by what other means. Are design and purpose something we can discuss about how our Universe came to being, as scientists, or should we keep a composed silence, and keep doing our reductionist things?
So, let us imagine, for the sake of arguing, that the big bang, and the consequent expansion of spacetime into what is today our Universe, is a common phenomenon. Well, for sure it must have a probability larger than zero of happening, according to some uber-laws of physics yet to be fathomed, or we would not be here to discuss it; so here "common" means "that may happen without violating any of those laws". It may be very rare, but it is clearly not impossible, so it happens by force of law. If it does, it will continue to happen forever. This, by itself, implies that an infinite number of Universes are bound to take place, and have taken place already at the time of speaking, if "time" really means anything in this context.
Another line of reasoning which leads to the idea of a multitude of Universes is the "multiverse" often broadcast as a necessary consequence of string theory. I do not know much about that, but here we can certainly take as a working hypothesis the idea of an infinite foam of bubbles, each one a different Universe with a different set of initial conditions and parameters.
Whatever the arrangement, we can make one further observation about our own Universe: that it is marvelously complex and rich of structure. Those properties stem from the precise value of some fundamental parameters. We do not know why the electron and proton masses have the values we observe, or the fine structure constant, or other defining parameters of Nature, but we certainly know that if those quantities had significantly different values then our Universe would be much less rich, and not lead to intelligent life (me).
But seriously, the marvelous complexity we see around may certainly be due to mere chance: the enormous selection bias of being here to witness that compexity is enough to explain whatever rarity you may assign to the emergence of a Universe which can host intelligent life at some point of its evolution.
But what if that complexity were instead something that the whole Universe-spewing machinery wanted to achieve by design, in a sort of trial-and-error process? We know that the parameters of Nature must be enormously well fine-tuned to give rise to such an interesting outcome. Thinking in terms of an optimization process, this might be pictured as a very carefully minimized loss function that some sort of automated machine-learning process produced, after a long and complex convergence procedure. Since there is an infinite time to achieve that convergence, even a very, very slow learning may do. Note that here I am referring to machine learning only because it is an easy analogy to what automatic minimization processes do by themselves.
The teleology of the construction is in imagining that there is something to optimize, or even in just asserting that there is a loss function of sorts, or anyway a converging process: this implies purpose. But if we buy into that kind of thinking, then we may even take the further step and ask ourselves: what kind of information can be learnt through physical processes from a Universe, from the outside of it? Because an iterative learning process requires that the output of some parameter choices may inform the choice of new parameters. This, by the way, is common to automatic minimization, too.
In the Universe-foam idea of the multiverse, for example, what can a bubble tell about itself, and the complexity and marvelous things that happened at some point in its interior, and how does it propagate it outside, such that some external system may "learn" that those parameters are better than others tried before?
These are no doubt mind-boggling questions, and maybe you could rightly call them silly. For sure they sit better in the realm of philosophy than in the one of physics, and so it is due time for me to return from my metaphoric walk -just as I am returning from my very real walk on daisies- and regain some composture and my role of die-hard reductionist and empirist. Yet maybe philosophy is indeed bound, at some point in the future, to re-unify with physics, after a journey of 2500 years that started with a few Greek thinkers who imagined atoms as the constituents of the world. Even leaving aside anthropic principles and similar lines of reasoning, the question remains: if our Universe is so special, is it a chance or some kind of evolutionary necessity? And if the latter is true, what is the mechanism? And what is the purpose?
Ai posteri l'ardua sentenza.