This chapter has two combined aims. First, I point out that the standard fine-tuning argument for the multiverse, as discussed in the previous two chapters, differs crucially from paradigmatic instances of anthropic reasoning such as, notably, Dicke's and Carter's accounts of large number coincidences between large numbers in cosmology. The key difference is that the standard fine-tuning argument for the multiverse treats the existence of forms of life as calling for a response and suggests to infer the existence of a multiverse as the best such response. Anthropic reasoning of the type championed by Dicke and Carter, in contrast, assumes the existence of forms of life as background knowledge when assessing whether the large number coincidences are to be expected, given the competing theories. The second aim of this chapter is to propose a new fine-tuning argument for the multiverse, which – unlike the standard one – is structurally similar to Dicke's and Carter's accounts of large number coincidences. The new argument turns out to have the virtue of being immune to the inverse gambler's fallacy charge.

The most-discussed objection against this argument is that it commits the inverse gambler's fallacy, originally identified by Ian Hacking. This fallacy consists in inferring from an event with a remarkable outcome that there have likely been many more events of the same type in the past, most with less remarkable outcomes. I discuss several suggested analogs to the problem of the fine-tuned parameters. Ultimately, as I argue, established standards of rationality may just not allow one to decide whether the standard fine-tuning argument for the multiverse commits the inverse gambler’s fallacy or not. Some of the considerations in this chapter, as explained along the way, are relevant to the debate about the Fermi paradox.

Multiverse theories are physical theories according to which we have empirical access only to a tiny part of reality that may not be representative of the whole. The idea that we might live in a multiverse is often suggested as a response to the finding that various parameters seem fine-tuned for life. The combination of string theory and inflationary cosmology is taken to propose a specific implementation of the multiverse idea: the landscape multiverse. I review these lines of thought and outline the structure of the book. I also highlight its central theses and ideas.

The self-sampling assumption can be seen as an indifference principle of self-locating belief: it instructs us to treat all the possibilities that might be in the reference class of observers as equally likely. Indifference principles of self-locating belief are regarded as suspect by some philosophers because they appear to have paradoxical consequences. Notably, an indifference principle of self-locating belief is usually appealed to in the notorious Doomsday Argument, and it also plays a role in the derivation of apparent “anomalous causal powers” in Nick Bostrom's Adam and Eve thought experiments. The recommendation that we should sometimes act as if there were anomalous causal powers seems very hard to accept. I show that reasoning akin to that used in the Doomsday Argument and in the Adam and Eve thought experiments leads to a similar recommendation in a version of the famous Sleeping Beauty problem. All these unattractive recommendations can be avoided if, as required by the BIC, one pays careful attention to the background evidence based on which one assigns probabilities to competing hypotheses and chooses the observer reference class in accordance with that background evidence.

This chapter continues the discussion of the standard fine-tuning argument for the multiverse, switching to the language of Bayesianism. After highlighting the desideratum of motivating a non-negligible (ur-) prior for the multiverse, I assess a worry, due to Cory Juhl, about belief in the multiverse, as based on the standard fine-tuning argument for the multiverse: that, even if the inverse gambler's fallacy charge could be rebutted, such belief would inevitably rely on illegitimate double-counting of the fine-tuning evidence. I argue that this concern can be assuaged, at least in principle: it is coherently possible for there be empirical evidence in favor of some specific multiverse theory – and thereby, derivatively, for the generalized multiverse hypothesis – whose evidential impact is independent of the fine-tuning considerations. The probabilistic formalism is also used to clarify why it is so difficult to determine whether the standard fine-tuning argument for the multiverse is fallacious: the difficulty can be linked to an ambiguity in the background knowledge based on which the impact of the finding that the conditions are right for life in our universe is assessed.

One of the most popular responses to the fine-tuning considerations is to suggest that parameters might have been set to life-friendly values by some cosmic designer or God. I discuss two important objections against this response: first, that we cannot infer anything from life-friendly parameters because we could not possibly have found ourselves in a lifeless universe (the anthropic objection) and, second, that there is no specific conception of the designer according to which the inference to a designer would be plausible: anthropomorphic conceptions of the designer are no longer credible; and on non-anthropomorphic conceptions, the designer’s preferences are inscrutable. I reject the first objection and endorse the second.

This chapter discusses further types of multiverse hypotheses: the branching worlds of Everettian quantum theory, the totality of possible worlds according to David Lewis’s modal realism, and the totality of mathematical structures, which – according to Max Tegmark – are all physically realized. I argue that suggested solutions to the probability problem in Everettian quantum theory are unconvincing. A recent proposal by Sebens and Carroll to derive the Born rule in the Everettian framework using considerations about self-locating belief turns out to suffer from a circularity problem because the central principle on which the derivation is based has no independent motivation besides, possibly, appeal to the Born rule itself. I further argue that it is not possible to coherently, in full seriousness, believe David Lewis's modal realism. Nor is it possible to coherently believe Max Tegmark's thesis that there is a multiverse of all mathematical structures in which they are physically realized. Tegmark's claim that our universe is itself a mathematical universe hypothesis is found to be incoherent.

According to many physicists, several aspects of the laws of nature, the constants, and the cosmic boundary conditions are fine-tuned for life: had they been slightly different, life would not have existed. Here I review the claimed instances of fine-tuning and some of the criticism that has been levelled against the fine-tuning considerations. I also discuss in which sense, if any, fine-tuned parameters may qualify as improbable. Finally, I review the naturalness criterion of theory choice and discuss how violations of naturalness may be regarded as relevant to the discussion about fine-tuning for life.

The chapter concludes the book with some reflections on the future of physics in the light of the possibility that we may neither ever be able to discard multiverse theories as pseudoscience nor obtain compelling evidence in favor of some multiverse theory or for the correct theory of fundamental physics as determinately “not” being a multiverse theory. We may never find out, in other words, whether there are other universes with different parameters. Our knowledge of physics “at large” – just as our knowledge of physics “at small” – may forever be limited.

Applying the BIC in practice is far from straightforward and fraught with difficulties because it requires the regularization of space-time infinities by implementing some cosmic “measure.” Furthermore, a suitable physical quantity must be chosen as proxy for the number of reference class observers in some given space-time region. Unfortunately, the choices made in this procedure are prone to being exploited – often unintentionally – by the researchers as so-called researcher degrees of freedom (a term from the social science literature) to yield those results that would best conform to their theoretical preferences. In the light of this difficulty, the prospects for obtaining compelling evidence in favor of any specific multiverse theory by testing whether our observations are those that typical multiverse inhabitants would make do look bad. As it turns out, the multiverse theories that have the best chances of being successfully tested empirically are those that do not behave as typical multiverse theories in important respects – i.e., those multiverse theories according to which all universes in the multiverse are similar or identical in a significant number of ways.