IB Philosophy Past Papers & Drilling (Paper 3, HL): Unseen Text Answers

Question 1 (reading and annotation: about 15 minutes)

[Unseen text – original TutorZone sample. Topic: philosophy and technology]
[1] Every evening a fifteen-year-old in Hong Kong watches short videos. She chooses them herself; nobody forces her. Twenty minutes later she is watching material she did not set out to watch, and by midnight she is tired, faintly ashamed, and unsure whether she wanted any of it. Her classmates report the same. The puzzle is not that the machine is powerful. It is that her wanting seems to have been settled somewhere she was not present.

[2] Call this the curation problem. A recommender does not merely offer options; it orders them, and it learns from every hesitation and every half-second of delay. It measures success by engagement, so its interest and the user’s long-term interest are not always the same. We say the system “knows” her; that is a metaphor. But the metaphor points at something quite concrete: a statistical model has formed a description of her dispositions finer-grained than any description she could give of herself.

[3] Two claims are commonly run together here. The first is causal: exposure shapes preference – nothing new, since families, schools, markets and friends have always done so. The second is normative: shaping someone’s preferences without her knowledge wrongs her. The first does not entail the second. Someone will object that education also shapes preference, and we do not therefore call education a wrong. So the real question is where influence ends and manipulation begins.

[4] A first answer appeals to transparency: influence is legitimate when its methods can be inspected. But transparency is neither necessary nor sufficient. A parent’s influence is rarely transparent to the child and is not therefore illegitimate; a disclosed dark pattern is still a dark pattern. Transparency matters, but it is not the whole story.

[5] A second answer appeals to autonomy. Autonomy is not having desires that nobody caused – that picture is unintelligible, since no one causes their own first desire. Autonomy is standing in a certain relation to one’s desires: being able to reflect on them, to endorse some and reject others, and to act on what one endorses. On this account the curation problem is not that the model manufactures desires, but that it works below the level at which reflection operates, and profits from discouraging reflection.

[6] That account has a cost. If autonomy requires reflective endorsement then autonomy comes in degrees rather than being all-or-nothing, and many ordinary lives – the exhausted parent, the grieving friend, the student in examination season – fall short of the ideal. A theory that makes autonomy the test of wrongdoing will find wrongdoing almost everywhere; that is a reason to doubt the test, not to doubt the world.

[7] A third answer appeals to relationality. Who we are is partly constituted by our relations: to language, to institutions, to the people who raised us. If so, the demand for a purely self-authored preference was never coherent, and the question becomes which relations are answerable to us and which are not. A feed that cannot be questioned, cannot explain itself, and is owned by a party whose interests diverge from ours, is a troubling relation – not because it influences us, but because it cannot be held to account.

[8] Notice what each answer presupposes. The transparency answer presupposes a public of informed agents; the autonomy answer presupposes a self divisible into a reflecting part and a part reflected upon; the relational answer presupposes that selves are partly made by their world. These are not empirical claims waiting to be settled by measurement. They are philosophical commitments, and the disagreement between them is not a disagreement about data.

[9] This is where philosophy earns its place in a technological argument. It will not tell an engineer how to write a ranking function. It does the slower work of making explicit what we already assume when we say someone has been wronged, of working out what those assumptions commit us to, and of asking which of them we can still defend when we are called to account. A society that skips that work will still decide; it will simply not know what it has decided.

[10] The student will probably never read any of this. She will, however, keep being asked to agree to terms she has not read, in a vocabulary built by others. Whether that counts as a wrong is not something the feed can settle.

Question 1 (part A: 10 marks; suggested 25 minutes)

Part A: Explain the claim in paragraphs [1] and [2] that “her wanting seems to have been settled somewhere she was not present”, and identify the conception of freedom it presupposes. Refer to the text and to your knowledge from the course. (10 marks)

Question 2 (part B: 15 marks; suggested 35 minutes)

Part B: Using the reasoning in paragraphs [8] and [9], discuss the nature and function of philosophy when it engages with contemporary technology, and the methods it uses. Refer to the text and to your own experience of doing philosophy in this course. (15 marks)