With the advent of modern AI techniques, especially the brain-inspired tools known as deep neural networks, scientists can now squeeze a huge amount of utility from datasets far too big to be analyzed by hand. In disciplines as diverse as neuroscience, biochemistry, and climate science, the marriage of big data and AI has yielded extraordinary practical advancements. But the deep neural networks that are enabling these astonishing practical advances offer few clues about the nature of the data they were trained on or the intellectual basis of their predictions: They perform end runs around human insight while making far more accurate and useful predictions than our scientific theories enable. The seeds that could one day enable science to transition into a wholly practical regime, in which human understanding is construed as an unnecessary indulgence, have now been sown.

The epilogue reflects on the unifying themes that emerge across the scientific ideas explored in this book and considers their implications for the future of knowledge. A central lesson is that modern science has revealed fundamental limits – to prediction, to certainty, and even to what can be known in principle. From Gödel’s incompleteness theorems and Turing’s limits of computation to quantum uncertainty and chaotic dynamics, uncertainty appears not as a failure of understanding but as a structural feature of reality. A second recurring theme is that extraordinary complexity can arise from simple underlying rules, whether in computation, physics, biology, or planetary processes. Looking forward, the epilogue considers the growing role of artificial intelligence in scientific discovery, suggesting that future insights may increasingly be generated by systems whose internal reasoning is difficult for humans to interpret, even when their results can be verified. The book concludes by arguing that making sense of modern science requires embracing both humility and curiosity, as humanity navigates a future shaped by powerful ideas, computational tools, and enduring uncertainty.

Throughout their history, philosophy and astronomy have been closely linked. Astronomy and cosmology inhabit novel epistemic spaces, and have novel epistemic aims, relative to more standard sciences. As the philosophy of astronomy emerges as an independent subdiscipline, the special nature of explanation and prediction within astronomy requires further examination and articulation. Unlike conventional experimental sciences, astronomy typically deals with entities and systems that are outside the reach of human intervention. Furthermore, astronomy and cosmology are historical sciences. This chapter evaluates the epistemology of astronomy by focusing on the canons of explanation, prediction, and theoretical success that are particular to astronomy. Demonstrations of the successes and shortcomings of philosophical models of explanation and prediction are given with respect to the eighteenth-century discovery of stellar aberration and the more recent conjectures of dark matter and dark energy.

Prediction is explored as both a core human cognitive function and a defining strength of AI systems. The chapter traces the history of prediction from statistical forecasting to modern personalization engines. It explains how humans rely on heuristics, experience, and context, while AI systems leverage large-scale data and probabilistic modeling. Applications in advertising, healthcare, and education illustrate AI’s predictive power, but the chapter also warns of pitfalls such as overfitting, bias, and Goodhart’s Law (when metrics distort outcomes). The key argument is that prediction succeeds when human judgment and machine learning complement one another, not when either acts alone.

This chapter argues that two of the common methods used in behavioural and social sciences to reduce the chances that models overfit the available data, namely heavy reliance on benchmark models and rigorous parameter estimation techniques, can slow the advancement of these sciences. An examination of classical decision research highlights how applying these methods shaped the field but have also led to limited success. As an alternative, the chapter proposes a prediction-oriented approach to the development of behavioural models. Evaluating and comparing models based on their predictive power inherently guards against overfitting and also facilitates accumulation of knowledge. The chapter reviews research employing the prediction-oriented approach in behavioural decision research and demonstrates that, in contrast to a common misconception, the focus on predictions can also facilitate better understanding of the underlying processes.

A theory of some finding or observation is an explanation of that finding or observation. Further, a good theory is a set of principles that are sufficient to show that the phenomenon is an instance of more general phenomena or principles. But not all explanations help us understand general phenomena because they lack some fundamental characteristics. The necessary characteristics of adequate explanations include explicit definitions and precise and limited scope, that is, they do not attempt to explain everything about a given event or action. Further, they can be tested with empirical data; they do not appeal to supernatural forces or to explanations with claims that testing is not necessary.

The case of bilingualism is a challenge for psycholinguists who aim to understand how the two (or more) languages of a bilingual are represented in the brain, whether they are organized similarly and how bilinguals manage to keep their languages apart. We first review studies that investigate the organization of the two languages in the brain and whether they interfere with each other during access to the lexicon and syntactic representations. In the second part of the chapter, we report neurolinguistic studies that examine cognitive processes and neural perspectives in monolinguals and bilinguals, with a special focus on factors that may influence bilingual language processing such as proficiency and age of acquisition. Finally, in the third part of the chapter we present theories on L2 processing and discuss the studies presented earlier in relation to these theories. In addition, we have extended the sections on lexical access in sentence context and syntactic processing by including recent studies that reflect the flourishing interest for bilinguals’ ability to predict upcoming words online during sentence comprehension.

In this chapter, we introduce the concept of regression, which is a way to quantify the relationship between an outcome Y and a vector of features X. This relationship is expressed by the regression function, which is the mean of Y given X. This chapter discusses the main ideas and goals of regression analysis.

Assembling and evaluating warrant for your claims involves giving reasons why these claims might be true or false. Such reasons may be understood as consisting of conjunctions of factual and relevance claims which may be adduced in favour of the claims in question. These claims themselves may in turn be warranted by further such conjunctions, and so on through indefinite higher ‘levels’ of evidence. We show how to structure relevant claims into an evidence role-map. We set out how to apply this to indirect local evidence, referencing a situation-specific causal equation model of each causal step. By reference to our pluralistic account of causation, we identify eight categories of evidence for each step. We introduce a notation for this approach and illustrate it using the Barbados sugar-sweetened beverage tax example, describing its use in post-hoc evaluation and ex ante prediction.

Music rhythm and speech rhythm share acoustic, temporal and syntactic similarities, and neuroscience research has shown that similar areas and networks in the brain are recruited to process both types of signals. Rhythm is a core predictive element for both music and speech, allowing for facilitated processing of upcoming, predicted elements. The combined study of music and speech rhythm processing can be particularly insightful, considering the stronger regularity and predictability of musical rhythm. Although speech rhythm is less regular, it still contains regularities, notably at syllabic and prosodic levels. In this chapter, we outline different research lines investigating connections between music and speech rhythm processing, including the recently proposed processing rhythm in speech and music framework, as well as music rhythm interventions and stimulations that aim to improve speech signal processing both in the short term and the long term. Implications for developmental language disorders and future research perspectives are outlined.

Spoken and written language are likely to share many aspects of how they are represented in the human mind. For instance, it would be highly inefficient for the brain to store the meaning of words separately for its spoken and written forms. Instead, shared representations across modalities allow for interaction between them, meaning that the effects of written language can directly influence spoken language processing. As a result, predictive learning that occurs during reading naturally transfers to spoken language. Knowledge accumulated through reading, along with the predictive behavior it fosters, can thus directly support prediction in speech as well.

Human fetuses, similar to adults, demonstrate a predictable response to a gradually evolving hypoxic stress due to a progressive increase in the frequency, duration, and strength of ongoing uterine contractions. Initial attempts to protect the myocardial workload to maintain the positive energy balance and preservation of aerobic metabolism within the myocardium is manifested by the onset of decelerations on the CTG trace. If the hypoxic stress progressively worsens, the fetus will release catecholamines to increase cardiac output and tissue perfusion, as well as to redistribute oxygenated blood from ‘non-essential’ organs by peripheral vasoconstriction to fetal central organs to ensure continuation of aerobic metabolism due to this ‘centralization’. Failure of these compensatory responses either due to poor fetal reserve, worsening intensity of hypoxia, or exhaustion of fetal reserves will lead to fetal decompensation and the onset of anaerobic metabolism within fetal central organs (i.e. brain, heart, and adrenal glands). These changes can be predicted by application of the knowledge of fetal physiological responses to a gradually evolving hypoxic stress.

This chapter discusses the move to modern meteorology, the science of weather. As meteorology has moved from antiquity through modernity, as we’ve sliced and diced the various aspects of weather into measurable, quantifiable units, we have demystified and changed our thinking about weather altogether. Without question, this conceptual slicing and dicing has increased our understanding of weather phenomena and improved the predictive validity of our forecasts, but it has also in many ways removed us from the most hazardous front lines of weather. The objective of this chapter is more epistemic than practical, to suggest that our relationship with weather has changed as we’ve learned to conceptualize weather differently. The final section of the chapter discusses the ways in which the demystification and quantification of weather has been adapted to characterize weather and its impacts as risk.