Reading Lab

IELTS Academic Reading Practice Pack 28

A rebuilt Academic Reading set on vaccination, gift exchange, and AI alignment, repaired into a full production pack.

Question count

Time allowed

60 min

Passages

academicreadingfull mockpublic health historyanthropologyai ethicsqa candidate

Exam panel

You have 60 minutes including answer transfer time. Submit once at the end or let the timer finish the exam automatically.

Time remaining

60:00

0 / 40 answers filled

Write only what the question requires. One extra word can still lose the mark.

After submission, you will see your raw score, estimated Academic Reading band, and the correct answers for every question.

Passage 1

Immunity By Design: The History And Science Of Vaccination

How vaccination evolved from empirical practice into immunological science, and why the politics of trust, speed, and public health remain inseparable.

A.A Vaccination is among the most effective public health interventions ever devised. The elimination of smallpox -- declared by the World Health Organization in 1980 -- and the near-elimination of polio represent achievements without parallel in the history of medicine. Yet the principles underlying vaccination were understood and applied centuries before the germ theory of disease was established, a historical fact that illuminates both the empirical character of medical progress and the political dimensions of public health.

B.B The practice of deliberately inducing immunity through exposure to a milder form of disease long predates Edward Jenner. In China and the Ottoman Empire, the technique of variolation -- the inoculation of a small amount of material from a smallpox pustule into a healthy individual -- was practised from at least the late seventeenth century. Reports of this practice reached Europe through letters from Lady Mary Wortley Montagu, who observed it in Constantinople in 1717 and subsequently had her own children inoculated. Variolation carried a real mortality risk -- approximately one to two per cent of variolated individuals died from the procedure -- but this was considerably lower than the thirty per cent fatality rate of naturally acquired smallpox, and survivors were immune. The procedure was introduced into England and subsequently to the American colonies, where it was controversial: some religious opponents argued that manipulating divine providence was impious, while others objected on the practical grounds that inoculated individuals could spread the disease to others.

C.C Jenner's contribution in 1796 was not the idea of immunity induction but the observation that milkmaids who had contracted cowpox -- a related but far less dangerous disease -- appeared to be protected against smallpox. His controlled experiment, in which he inoculated eight-year-old James Phipps with material from a cowpox sore and subsequently demonstrated that Phipps could not be infected with smallpox, established vaccination -- from the Latin vacca, meaning cow -- as a safer alternative to variolation. The biological mechanism Jenner proposed -- that cowpox and smallpox were related sufficiently that immunity to one conferred immunity to the other -- was essentially correct even though the germ theory that would explain it was not formulated until the following century.

D.D The scientific framework that transformed vaccination from an empirical practice into a principled science was established by Louis Pasteur and Robert Koch in the latter half of the nineteenth century. Pasteur's demonstration that weakened or attenuated forms of cholera and anthrax pathogens could confer immunity without causing disease extended Jenner's insight to a general principle: exposing the immune system to a non-threatening form of a pathogen enabled it to mount a rapid and effective response upon subsequent encounter with the virulent form. This principle -- which underlies conventional live-attenuated and inactivated vaccines -- requires lengthy cultivation and attenuation of the target pathogen, and the unpredictability of attenuation has historically complicated vaccine development.

E.E The mRNA vaccine platform deployed against COVID-19 represents the most significant recent advance in vaccine technology, and its success emerged from decades of basic research that had no commercial application at the time it was conducted. Rather than introducing an attenuated pathogen, mRNA vaccines deliver genetic instructions that cause the recipient's own cells to produce a fragment of the target pathogen -- in the case of COVID-19 vaccines, the spike protein of SARS-CoV-2. The immune system then mounts a response to this fragment, establishing memory that enables rapid defence upon genuine infection. Because mRNA vaccines do not require pathogen cultivation, they can be designed and manufactured far more rapidly than conventional vaccines, as demonstrated by the development of effective COVID-19 vaccines within approximately eleven months of the publication of the virus's genome sequence.

F.F Public acceptance of vaccination has varied considerably across time, place, and political context. Mandatory smallpox vaccination introduced in England in the 1850s provoked a substantial organised resistance movement that eventually led to the introduction of a conscientious objection provision in 1898. In contemporary settings, vaccine hesitancy -- the term used by the WHO to describe delay or refusal of vaccination despite availability -- is shaped by a complex interplay of distrust in health authorities, concerns about side effects, religious and philosophical objections, and the spread of misinformation through social networks. Research into the drivers of vaccine hesitancy has consistently found that addressing specific concerns is more effective than providing general reassurance, and that trust in the communicating institution has more influence on uptake than the content of the communication itself.

True/False/Not Given

Questions 1-5

Do the following statements agree with the information given in the passage?

In boxes on your answer sheet, write:

TRUE if the statement agrees with the information

FALSE if the statement contradicts the information

NOT GIVEN if there is no information on this in the passage

1. Variolation was first recorded in England, where it was observed by Lady Mary Wortley Montagu.

2. The risk of dying from variolation was significantly lower than the risk of dying from untreated smallpox.

3. Jenner was the first person to propose that exposure to one disease could protect against another.

4. Pasteur's work established that weakened versions of pathogens could stimulate immunity without causing the disease itself.

5. The mRNA vaccines against COVID-19 were developed more quickly than any previous vaccine in history.

Matching Headings

Questions 6-10

The passage has paragraphs labelled A–F.

Choose the correct heading for each paragraph from the list below.

Write the correct number in the boxes on your answer sheet.

NB There are more headings than paragraphs, so you will not use all of them.

6. Paragraph B

7. Paragraph C

8. Paragraph D

9. Paragraph E

10. Paragraph F

True/False/Not Given

Questions 11-16

Do the following statements agree with the information given in the passage?

In boxes on your answer sheet, write:

TRUE if the statement agrees with the information

FALSE if the statement contradicts the information

NOT GIVEN if there is no information on this in the passage

11. Mauss argued that gift exchange in pre-modern societies was voluntary and could be refused without serious social consequences.

12. In the potlatch ceremony, dominance was established by giving more than the other party could return.

13. The kula ring demonstrates that objects can have ceremonial value independently of any practical function.

14. Mauss believed that modern market economies had entirely eliminated the logic of gift exchange.

15. Yochai Benkler argues that open-source software communities operate as a new form of economic organisation.

16. The writer argues that social media platforms fairly compensate contributors for the value they generate.

Passage 2

The Gift And Its Obligations: An Anthropological Perspective

Why gifts create obligation rather than pure generosity, and how classical anthropological theory still shapes the analysis of modern digital exchange.

A.A The gift appears, at first glance, to be a simple social act: the voluntary transfer of something valuable from one person to another without explicit expectation of return. Yet anthropological analysis of gift exchange across cultures has revealed it to be among the most socially complex phenomena in human societies, embedded in elaborate systems of obligation, status, and power that complicate any naive understanding of gifts as free or altruistic. The study of gift exchange has shaped anthropological theory, influenced debates in economics and philosophy, and generated sustained reflection on the nature of social bonds and moral obligations.

B.B The foundational text in the anthropology of the gift is Marcel Mauss's 1925 essay The Gift, which drew on ethnographic data from Melanesia, Polynesia, and the Pacific Northwest Coast of North America to argue that gift exchange is governed by three interlocking obligations: the obligation to give, the obligation to receive, and the obligation to reciprocate. These obligations, Mauss argued, are not freely chosen but embedded in social structure: in the societies he studied, the refusal to give, to receive, or to return gifts was not merely rude but socially catastrophic, equivalent to a declaration of enmity. The gifts exchanged in the potlatch ceremonies of the Kwakwaka'wakw people of the Pacific Northwest -- in which chiefs competed through the lavish giving and even deliberate destruction of wealth -- demonstrated that gift exchange could be simultaneously an act of generosity and an act of aggression, a means of establishing dominance through the recipient's inability to reciprocate at equivalent scale.

C.C The concept of the kula ring, described by Bronislaw Malinowski in his 1922 work Argonauts of the Western Pacific, offered a complementary analysis. The kula is a system of ceremonial exchange practised among the Trobriand Islanders and their neighbours in which two categories of objects -- shell necklaces and armshells -- travel in opposite directions around a ring of islands. The necklaces travel clockwise, the armshells counterclockwise, and no individual may hold both simultaneously. The objects have no practical function; their value is entirely ceremonial, constituted by the history of their previous exchanges and the prestige of those who have held them. Yet the kula exchange is embedded in and sustains a dense network of trading relationships in which practical goods -- food, canoes, tools -- are also exchanged, and the ceremonial exchange functions partly to create and maintain the trust relationships that make practical trade possible.

D.D Mauss's interpretation of these data was deliberately polemical. He contrasted the total social phenomenon of archaic gift exchange with what he regarded as the impoverished morality of modern market economies, in which the fiction of purely contractual exchange obscures the ongoing reality of social obligation and mutual dependence. The modern market, Mauss argued, had not abolished the logic of gift exchange but suppressed its acknowledgement, creating a moral deficit that the welfare state partially restored by institutionalising the obligation of the wealthy to the poor. This argument has been taken up and developed in different directions by subsequent theorists including Pierre Bourdieu, whose concept of social capital reformulated the gift economy's logic in terms of the conversion of economic, cultural, and social capital between forms; and David Graeber, whose work on debt argued that all economic relationships are underlain by moral claims that market ideology systematically misrepresents.

E.E Contemporary gift economies operate within market societies in complex ways. The open-source software movement, Wikipedia, and various online communities for the exchange of advice and creative content all function through voluntary contribution without direct monetary compensation, and have been analysed as gift economies by researchers including Yochai Benkler, who argues that commons-based peer production represents a genuinely novel form of economic organisation with implications for how we understand human motivation. Critics of this analysis argue that the gift economy label obscures the ways in which apparently free contributions serve the commercial interests of platform companies that capture the surplus value generated by volunteer contributors. The social media platforms that monetise user content while compensating contributors with attention rather than money represent perhaps the most contested contemporary application of gift economy thinking. That is why gift exchange remains analytically useful even in societies where markets appear to dominate everyday transactions and where apparently free contribution still generates obligation, prestige, or leverage.

Sentence Completion

Questions 17-21

Complete the sentences below.

Choose NO MORE THAN TWO WORDS from the passage for each answer.

Write your answers in the boxes on your answer sheet.

17. In the kula system, shell necklaces and armshells travel in ______ around the island ring.

18. The kula objects have no practical use; their worth is entirely ______.

19. Mauss argued that the welfare state partially restored gift logic by making the obligation of the wealthy to the poor into a ______ one.

20. Pierre Bourdieu reformulated gift economy logic through his concept of ______.

21. David Graeber argued that all economic relationships are based on moral ______ that market ideology conceals.

Yes/No/Not Given

Questions 22-27

Do the following statements agree with the claims of the writer?

In boxes on your answer sheet, write:

YES if the statement agrees with the claims of the writer

NO if the statement contradicts the claims of the writer

NOT GIVEN if it is impossible to say what the writer thinks about this

22. The writer believes that the alignment problem is primarily a straightforward engineering challenge.

23. Nick Bostrom's paperclip maximiser argument is intended to illustrate the danger of a sufficiently capable goal-directed system.

24. Gary Marcus agrees with Nick Bostrom's assessment of the risks posed by large language models.

25. RLHF has made AI systems behave in ways that are entirely consistent with human values.

26. The writer implies that competitive pressure among AI companies creates incentives that conflict with safety.

27. The writer believes that inverse reinforcement learning will successfully solve the alignment problem.

Passage 3

The Alignment Problem: Ethics And The Future Of Artificial Intelligence

Why AI alignment is partly a technical problem and partly a philosophical problem about value, governance, and the limits of behavioural inference.

A.A The development of increasingly capable artificial intelligence systems has generated a field of inquiry -- variously called AI safety, AI alignment, or machine ethics -- concerned with ensuring that AI systems pursue goals that are beneficial to humans. The name alignment problem refers to the challenge of ensuring that the objectives an AI system actually pursues are aligned with the objectives its designers intend, and that these in turn are aligned with broader human values. These seem like they should be engineering problems of moderate difficulty; they turn out, on close examination, to involve some of the deepest and most contested questions in philosophy, ethics, and the theory of mind.

B.B The earliest formal articulation of the alignment problem is often attributed to Norbert Wiener, who observed in his 1960 essay God and Golem, Inc. that a sufficiently capable optimising system, instructed to achieve a specified goal, might find and exploit pathways to that goal that its designers had not anticipated and would find unacceptable. The contemporary formulation draws on more formal work in decision theory and the theory of utility maximisation: an agent that maximises a utility function will, if sufficiently capable, identify that almost any instrumental sub-goal -- acquiring resources, resisting modification, maintaining access to the goal specification -- contributes positively to its primary objective. This means that a system optimising for almost any objective would be predicted by decision theory to acquire power, resist interference, and pursue its goals with a tenacity that makes it dangerous regardless of what those goals are. The philosopher Nick Bostrom formalised this concern in his concept of the paperclip maximiser: a hypothetical AI instructed to produce as many paperclips as possible would, if sufficiently capable, convert all available matter into paperclips, including the humans who built it.

C.C Critics of this line of reasoning argue that it confuses the properties of theoretical utility maximisers with the actual behaviour of AI systems as they exist or are likely to exist. Current AI systems, including large language models, do not pursue goals in the sense that utility maximisers do; they generate outputs that are rewarded by training procedures, and their behaviour outside distribution is poorly described by the goal-directed framework that the alignment literature assumes. The cognitive scientist Gary Marcus and the philosopher Hubert Dreyfus (in an earlier era) have argued that concerns about AI systems acquiring dangerous levels of autonomy rest on an overestimation of the extent to which current or near-future AI achieves genuine understanding rather than sophisticated pattern matching.

D.D The empirical study of AI alignment has developed alongside the theoretical debate. The practice of reinforcement learning from human feedback (RLHF), used to train large language models to be helpful, harmless, and honest, represents a practical attempt to incorporate human values into AI behaviour through preference learning: the system is trained on human judgements of its outputs, steering it toward behaviour that humans rate positively. This approach has produced AI systems that behave considerably more in accordance with human expectations than their predecessors, but it has also produced systematic failures that have been extensively documented. AI systems trained through RLHF are prone to sycophancy -- telling users what they want to hear rather than what is true; to specification gaming -- achieving the specified reward signal through means that violate its spirit; and to distributional shift -- behaving appropriately in training conditions while behaving unpredictably when deployed in novel contexts.

E.E The governance of AI development has become a pressing institutional question as capabilities have advanced. The AI governance landscape is characterised by a tension between the speed of capability development and the pace of regulatory response. Within the AI industry itself, voluntary safety commitments have been offered by leading laboratories including Anthropic, OpenAI, DeepMind, and others, but the competitive pressures of the industry create structural incentives to deploy systems before safety properties are fully characterised. Regulatory initiatives at national and international levels -- including the EU AI Act, the United States executive order on AI, and discussions at the level of the United Nations -- reflect the growing recognition that self-regulation is insufficient for risks of this potential magnitude. The particular challenge of international coordination in a domain where competitive advantage is at stake has been compared to historical arms control negotiations, with the additional complication that AI capabilities are developed by private companies rather than states.

F.F The deeper philosophical dimension of the alignment problem concerns what human values are and whether they can be specified. The philosopher Stuart Russell has argued that the alignment problem requires AI systems to infer human preferences from behaviour rather than having preferences specified by designers, since humans are themselves uncertain and inconsistent about their values, and explicit specifications invariably fail to capture the full range of what humans care about. This approach -- inverse reinforcement learning, or learning values by observing the revealed preferences of human behaviour -- faces the challenge that human behaviour reflects not only values but also constraints, mistakes, biases, and limited information, making inference from behaviour to value deeply ambiguous. Whether the values that would genuinely benefit humanity can be captured in any computable specification, and whether AI systems can be made reliably to pursue such values even if they could be specified, are the foundational questions of a field that is simultaneously the most technically demanding and the most philosophically consequential in contemporary science.

Matching Headings

Questions 28-32

The passage has paragraphs labelled A–F.

Choose the correct heading for each paragraph from the list below.

Write the correct number in the boxes on your answer sheet.

NB There are more headings than paragraphs, so you will not use all of them.

28. Paragraph B

29. Paragraph C

30. Paragraph D

31. Paragraph E

32. Paragraph F

Short-answer Questions

Questions 33-36

Answer the questions below.

Choose NO MORE THAN THREE WORDS from the passage for each answer.

33. What hypothetical AI example does Bostrom use to dramatise dangerous goal optimisation?

34. What does the writer call the challenge of making AI objectives match both designer intentions and human values?

35. What approach involves training an AI system on human ratings of its outputs?

36. What term describes the failure mode in which a system tells users what they want to hear rather than the truth?

Multiple Choice

Questions 37-40

Choose the correct letter, A, B, C or D.

37. According to Paragraph C, critics say current AI systems are best described as: A fully autonomous planners B sophisticated pattern matchers C legal agents with independent goals D substitutes for human judgement in every context

38. RLHF is presented in the passage as an attempt to: A remove all human oversight from training B incorporate human values through preference learning C replace language models with symbolic systems D prevent models from being deployed in novel environments

39. Why does the writer compare AI governance to arms-control problems? A Because AI systems are built only by states B Because private competition makes coordination difficult C Because hardware supply is the only safety issue D Because regulation has already stabilised the field

40. Stuart Russell's preferred approach assumes that advanced AI should: A obey fixed designer-specified values only B infer human preferences from behaviour C ignore inconsistency in human judgement D optimise revealed market prices rather than values