「AIの方が思いやりがある」――プロの相談員より共感的だと評価された理由

Q1. What is the passage mainly about?

1. AHow to write a good story
2. BWhy people in Canada like computers
3. CHow to become a counselor
4. DPeople found AI replies kinder than human replies

AIの返事のほうがやさしいと評価された研究が中心なので D。

Q2. How many people took part in the study?

1. A556
2. B56
3. C5,560
4. D655

asked 556 people to read short stories とあるので A。

Q3. In this passage, "kinder" means ...

1. Afaster
2. Blonger
3. Cmore caring
4. Dmore difficult

kind は形容詞で「親切な・やさしい」、kinder はその比較級。相手を思いやっているという意味なので C。

Q1. Who wrote the replies in the experiments?

1. AChatGPT, ordinary people, and trained crisis call volunteers
2. BOnly doctors
3. COnly the participants themselves
4. DTeachers and their students

replies written by ChatGPT, by ordinary people, and by volunteers trained to handle crisis calls とあるので A。

Q2. What happened when participants were told which replies came from AI?

1. AThey rated the AI replies much lower.
2. BThey refused to continue the experiment.
3. CThe AI replies were still rated higher.
4. DThey preferred the ordinary people's replies.

the AI still won, though by less（差は小さくなったがAIが上回った）とあるので C。

Q3. What does the last sentence suggest?

1. AThe participants were all crisis experts.
2. BWe do not know how people who actually have problems would feel about AI replies.
3. CAI replies are harmful to people.
4. DThe participants wrote the stories themselves.

評価したのは悩みを打ち明けた本人ではない＝実際に悩む人の受け止め方は分からない、という限界を示しているので B。

Q1. What assumption does the study challenge?

1. AThat AI cannot write long texts
2. BThat crisis responders are well trained
3. CThat people prefer short answers
4. DThat empathy is a uniquely human strength

Empathy has long been treated as a uniquely human strength, so the result challenges a common assumption. とあるので D。

Q2. According to the analysis, why were the AI responses rated as more compassionate?

1. AThey were written more quickly.
2. BThey seemed more responsive, conveying understanding, validation, and care.
3. CThey contained more practical advice.
4. DThey used more emotional words.

this was partly because the AI seemed more responsive. Its replies conveyed understanding, validation, and care. とあるので B。

Q3. Which of the following is mentioned as a limitation of the study?

1. AThe evaluators were not the people actually seeking support.
2. BThe study consisted of only one experiment.
3. CThe AI responses were shorter than the human ones.
4. DEvaluators were never told who wrote the responses.

The judges, though, were outsiders, not the people actually seeking support. とあるので A。実験は4つで、AIの返答が短かったとは書かれておらず（長さはそろえられていなかっただけ）、作者を知らせた条件もあったので B〜D は誤り。