Audits of multilingual resources are reporting shockingly poor quality: “less than 50% … acceptable quality.” There is too much translationese in too many of our multilingual resources, e.g., Wikipedia, XNLI, FLORES, WordNet. We view translationese as a form of noise that makes it hard to generalize from a benchmark based on translation to a real task of interest that does not involve translation. Worse, too much of this translationese is in the “wrong” direction. Directionality matters. Professional translators translate from their weaker language into their stronger language. Unfortunately, many of our resources translate in the other direction, from a stronger (higher-resource) language into a weaker (lower-resource) language. In Wikipedia, for example, there is more translation out of English than into English. We recommend more investments in high-quality data, and less in translation, especially in the “wrong” direction.

Retrieval-augmented generation (RAG) adds a simple but powerful feature to chatbots, the ability to upload files just-in-time. Chatbots are trained on large quantities of public data. The ability to upload files just-in-time makes it possible to reduce hallucinations by filling in gaps in the knowledge base that go beyond the public training data such as private data and recent events. For example, in a customer service scenario, with RAG, we can upload your private bill and then the bot can discuss questions about your bill as opposed to generic FAQ questions about bills in general. This tutorial will show how to upload files and generate responses to prompts; see https://github.com/kwchurch/RAG for multiple solutions based on tools from OpenAI, LangChain, HuggingFace transformers and VecML.

Astrobiology is a scientific endeavour involving great uncertainties. This could justify intellectual risk-taking associated with research that significantly deviates from the mainstream, to explore new avenues. However, little is known regarding the effect of such maverick endeavours. To better understand the need for more or less risk in astrobiology, we investigate to what extent high-risk / high-impact research contributes to breakthrough results in the discipline. We gathered a sample of the most impactful astrobiology papers of the past 20 years and explored the degree of risk of the research projects behind these papers via contact with the corresponding authors. We carried out interviews to explore how attitudes towards risk have played out in their work, and to ascertain their opinions on risk-taking in astrobiology. We show the majority of the selected breakthrough results derive from endeavours considered medium- or high-risk, risk is significantly correlated with impact, and most of the discussed projects adopt exploratory approaches. Overall, the researchers display a distribution of attitudes towards risk from the more cautious to the more audacious, and are divided on the need for more risk-taking in astrobiology. Our findings ultimately support the explicit implementation of a risk-balanced portfolio in astrobiology.

On p. 10 of the 2018 National Academies Exoplanet Science Strategy document (NASEM 2018), ‘Expect the unexpected’ is described as a general principle of the exoplanet field. But for the next 150 pages, this principle is apparently forgotten, as strategy decisions are repeatedly put forward based on our expectations. This paper explores what exactly it might mean to ‘expect the unexpected’, and how this could possibly be achieved by the space science community. An analogy with financial investment strategies is considered, where a balanced portfolio of low/medium/high-risk investments is recommended. Whilst this kind of strategy would certainly be advisable in many scientific contexts (past and present), in certain contexts – especially exploratory science – a significant disanalogy needs to be factored in: financial investors cannot choose low-risk high-reward investments, but sometimes scientists can. The existence of low-risk high-impact projects in cutting-edge space science significantly reduces the warrant for investing in high-risk projects, at least in the short term. However, high-risk proposals need to be fairly judged alongside medium- and low-risk proposals, factoring in both the degree of possible reward and the expected cost of the project. Attitudes towards high-risk high-impact projects within NASA since 2009 are critically analysed.

Ernietta plateauensis Pflug, 1966 is the type species of the Erniettomorpha, an extinct clade of Ediacaran life. It was likely a gregarious, partially infaunal organism. Despite its ecological and taxonomic significance, there has not been an in-depth systematic description in the literature since the original description fell out of use. A newly discovered field site on Farm Aar in southern Namibia has yielded dozens of specimens buried in original life position. Mudstone and sandstone features associated with the fossils indicate that organisms were buried while still exposed to the water column rather than deposited in a flow event. Ernietta plateauensis was a sac-shaped erniettomorph with a body wall constructed from a double layer of tubes. It possessed an equatorial seam lying perpendicular to the tubes. The body is asymmetrical on either side of this seam. The tubes change direction along the body length and appear to be constricted together in the dorsal part of the organism.

Rangea is the type genus of the Rangeomorpha, an extinct clade near the base of the evolutionary tree of large, complex organisms which prospered during the late Neoproterozoic. It represents an iconic Ediacaran taxon, but the relatively few specimens previously known significantly hindered an accurate reconstruction. Discovery of more than 100 specimens of Rangea in two gutter casts recovered from Farm Aar in southern Namibia significantly expands this data set, and the well preserved internal and external features on these specimens permit new interpretations of Rangea morphology and lifestyle. Internal structures of Rangea consist of a hexaradial axial bulb that passes into an axial stalk extending the length of the fossil. The axial bulb is typically filled with sediment, which becomes increasingly loosely packed and porous distally, with the end of the stalk typically preserved as an empty, cylindrical cone. This length of the axial structure forms the structural foundation for six vanes arranged radially around the axis, with each vane consisting of a bilaminar sheet composed of a repetitive pattern of elements exhibiting at least three orders of self-similar branching. Rangea was probably an epibenthic frond that rested upright on the sea bottom, and all known fossil specimens were transported prior to their final burial in storm deposits.

For many years—and with some energy since Laudan’s “Confutation of Convergent Realism” (1981)—the scientific realist has sought to accommodate examples of false-yet-successful theories in the history of science. One of the most prominent strategies is to identify ‘success fueling’ components of false theories that themselves are at least approximately true (judging by our current understanding). In this article I develop both sides of the debate, introducing new challenges from the history of science as well as suggesting adjustments to the divide et impera realist strategy. A new ‘recipe’ for the prospective identification of (at least some) working/idle posits is considered.

Research in unconventional medicine requires a number of different questions to build up a “mosaic” of evidence. Choice of research design depends on the question being asked and is independent of the therapy under investigation. Despite the doubts of some practitioners, randomized trials are of value for determining certain questions in alternative medicine.

Increasing evidence suggests a role for prenatal environment in the onset of cardiovascular and metabolic disease in later life. In the rat, undernutrition in utero and a postnatal high-fat diet gives rise to a phenotype similar to the metabolic syndrome. As endothelial dysfunction is a feature of both CVD and the metabolic syndrome we investigated the impact of maternal undernutrition and/or postnatal high-fat on endothelial function. Virgin Wistar rats were mated and randomly assigned to groups to receive food either ad libitum (control) or at 30 % of ad libitum intake throughout gestation. At postnatal day 250, a cohort from each group was challenged with a high-fat diet (D12451, 45 % energy from fat; Research Diets, Inc., New Brunswick, NJ, USA) for the remainder of the study. At 1 year of age, small mesenteric arteries were dissected and mounted on a wire myograph and responses to phenylephrine, endothelin, acetylcholine, leptin and sodium nitroprusside assessed. Vasoconstriction to endothelin was significantly enhanced in all groups compared with controls (-log effective concentration equal to 50 % of the maximal response (pEC50); P < 0·001). Endothelium-dependent vasodilatation to acetylcholine was significantly blunted in all groups compared with controls (% maximum response; P < 0·01), while dilatation to leptin and sodium nitroprusside was similar in all groups. These data demonstrate that both maternal undernutrition and postnatal high fat lead to vascular alterations and suggest that maternal undernutrition alone is at least as detrimental to offspring endothelial function as a long-term exposure to a high-fat diet in the offspring.

Obesity and type 2 diabetes are serious health issues in the developed world and are becoming increasingly important on a global scale. Furthermore, the marked increases in both childhood obesity and type 2 diabetes will translate to further increases in adult obesity, diabetes and associated co-morbidities in the near future; as such it has been ranked as a critical public health threat. It is a widely held view that the primary cause of obesity is the development of an obesogenic environment, due to ease of access to highly calorific food and reduced energy expenditure in work and leisure activities. In addition there is strong evidence for a genetic component to human obesity with the identification of a number of genes associated with human obesity. However, on its own the genetic component of this condition cannot account for the dramatic increase in the prevalence of obesity in recent years. Of relevance and as highlighted by epidemiological and experimental studies, is the relationship between the periconceptual, fetal and early infant phases of life and the subsequent development of adult obesity. The terms “developmental programming” and the “Developmental Origins of Adult Health and Disease” are preferentially used to describe these relationships. Despite initial controversy when these relationships were first suggested, both prospective clinical and experimental studies have clearly shown that the propensity to develop abnormalities of cardiovascular, endocrine and metabolic homeostasis in adulthood are increased when fetal development has been adversely affected. This pathogenesis is not based on genetic defects but on altered gene expression seen as a result of fetal adaptation to an adverse intrauterine environment. The relative role of genetic versus environmental factors and the mechanisms underlying developmental programming remain speculative. It is generally argued that in response to an adverse intrauterine environment, the fetus adapts its physiological development to maximise its immediate chances for survival. Owing to the plasticity of the fetus, these adaptations may include resetting of metabolic homeostasis and endocrine systems and the down-regulation of growth, commonly reflected in an altered birth phenotype. It is thought that whilst these changes in fetal physiology (i.e. the prenatal environment) may be beneficial for short term survival in utero they may be maladaptive in postnatal life, contributing to poor health outcomes when offspring are exposed to catch-up growth, diet-induced obesity and other factors. The “predictive adaptive response” hypothesis proposes that the degree of mismatch between the pre- and postnatal environments is a major determinant of subsequent disease. This review will address recent work in animal models and observations in the clinical and epidemiological settings on in utero adaptations and subsequent development of obesity and type 2 diabetes.

The needs of the researcher in the fine arts are considered. The nature of inter-lending in the field is outlined, and in particular, the role of the British Library Lending Division and its ability to supply this material are examined, together with an appraisal of its current stocks and acquisitions policy.

The opinions expressed in this article are those of the authors and do not necessarily reflect the view of the British Library.