Updated 2026-08-28hard

Duolingo English Test Interactive Listening Practice #12 — Hard Level

Free DET (Duolingo English Test) Interactive Listening practice #12 — Hard level. Just DET Listening questions, no grading — solve it yourself and check the answer or model response.

How to practice: This set collects only the Interactive Listening question type from the real DET (Duolingo English Test) — DET Listening questions. No grading, no timer — solve it yourself, then tap each question to reveal the answer or model response.
Conversation 1

Two ethicists are debating whether artificial intelligence systems can be genuine moral agents.

AThe key question is whether moral agency requires consciousness, or whether a sufficiently sophisticated behavioral profile — consistent rule-following, sensitivity to context, adaptive response to novel situations — is sufficient.
BI think that framing concedes too much. Moral agency isn't just about behavioral outputs. It requires that the agent understand why the rules matter — that they grasp the values behind the constraints, not just execute them.
ABut how do we verify that kind of understanding in any agent, human or artificial? We infer moral understanding from behavior and self-report, and we have no direct access to another's inner life.
BThat's a genuine epistemic problem, but it cuts differently for AI. With humans, we have evolutionary, neurological, and developmental grounds for attributing genuine moral comprehension. With current AI systems, we have none of those.
ASo your position is that the attribution of moral agency is currently unwarranted for AI, but the door isn't necessarily closed in principle?
Q1 · Interactive Listening
What does B say next?
  1. A. That's a fair characterization. If we develop systems that possess the relevant properties — whatever those ultimately turn out to be — the attribution may become warranted. But we're not there yet, and pretending otherwise has real ethical risks.
  2. B. No, I think AI will never be capable of genuine moral reasoning under any circumstances.
  3. C. You're right that behavior is sufficient for moral agency — we should grant it to current systems.
  4. D. The question of AI consciousness is irrelevant to the practical deployment of these systems.
Show Answer
Correct: That's a fair characterization. If we develop systems that possess the relevant properties — whatever those ultimately turn out to be — the attribution may become warranted. But we're not there yet, and pretending otherwise has real ethical risks.
Conversation 2

Two professors of philosophy are discussing whether the trolley problem has genuine implications for real-world ethical decision-making.

AMy concern with the trolley problem is that it's so stripped of context that the intuitions it elicits may not generalize to real moral situations, which always involve uncertainty, ongoing relationships, and institutional constraints.
BI take that point seriously, but I'd argue the abstraction is precisely the value. It isolates variables that are conflated in real cases, allowing us to examine one moral tension at a time without the noise of everything else.
AExcept that the noise — the uncertainty, the relational context, the institutional pressures — is exactly what real moral decision-making has to navigate. Training intuitions on idealized cases may actually mislead people.
BPerhaps. But would you want surgeons or military commanders making life-and-death decisions without ever having thought carefully about the structure of tragic choice? The trolley problem is a tool for developing that capacity, not a simulation of reality.
AI grant that. My worry is less about the thought experiment itself and more about how it gets used — as if resolving it settled the real ethical questions rather than sharpening the conceptual tools for approaching them.
Q2 · Interactive Listening
What does B say next?
  1. A. That's a criticism I can accept. The problem lies not in the method but in overstating its conclusions — it illuminates a structure without resolving the question of which structure applies to any given real-world case.
  2. B. I think the trolley problem is a perfect model for all real-world ethical situations.
  3. C. You're right that we should stop teaching the trolley problem in philosophy courses.
  4. D. Real-world ethical decisions are simpler than the trolley problem, so the comparison is misleading.
Show Answer
Correct: That's a criticism I can accept. The problem lies not in the method but in overstating its conclusions — it illuminates a structure without resolving the question of which structure applies to any given real-world case.
Conversation 3

A journalist is interviewing a bioethicist about the ethical implications of emerging genetic technologies.

AGermline editing — modifying DNA in embryos in ways that can be passed to future generations — seems qualitatively different from other medical interventions. Can you articulate why?
BIt crosses what I'd call the consent horizon. Every other medical intervention, however invasive, is performed on an agent who has the capacity to consent or refuse, either directly or through a proxy. Germline editing affects individuals who do not yet exist and cannot be consulted.
ABut that applies equally to vaccination programs, environmental policies, or any decision that affects future generations. Do those also violate an ethical principle?
BThere's an important asymmetry. Those decisions affect the conditions in which future people will live their lives. Germline editing alters the biological identity of the future person themselves. We are not just shaping their world; we are constituting them.
ASome would say that if the modification eliminates a heritable disease, the future person would endorse it if they could.
Q3 · Interactive Listening
What does B say next?
  1. A. That argument has intuitive appeal, but it proves too much. If we can justify any modification a future person 'would endorse,' we have licensed an extraordinarily broad set of interventions based on our own projections of someone who doesn't yet have preferences to consult.
  2. B. In that case, germline editing is fully justified and should proceed without restriction.
  3. C. Future persons have no morally relevant interests, so the consent issue doesn't apply here.
  4. D. Medical professionals alone should make all decisions about germline editing without public input.
Show Answer
Correct: That argument has intuitive appeal, but it proves too much. If we can justify any modification a future person 'would endorse,' we have licensed an extraordinarily broad set of interventions based on our own projections of someone who doesn't yet have preferences to consult.

More “Interactive Listening” sets at other levels · Browse the full DET question bank