登录
Passage 12
People are talking a lot about artificial intelligence(AI), viewing it as a force that could reshape how society works. But there is something important missing from this discussion. It isn't enough to ask how it will change us. We also need to understand how we shape AI and what it can tell us about ourselves.
Every AI model we develop mirrors our rules and expresses our beliefs. A few years ago, while looking for new workers, a famous company gave up an AI-powered tool after finding it unfavourable to women. The AI was not designed to behave this way. Instead, it was influenced by historical data favouring men. Similarly, a recent study found that lending algorithms(算法) often offer less favourable terms to people of colour, worsening long-standing unfairness in the lending industry. In both cases, AI isn't creating new biases(偏见); it is mirroring the ones that are already present.
These reflections(反映) give us an important chance to take a close look at ourselves. By making these problems seen and more pressing, AI challenges us to recognize and address what causes algorithmic bias. As AI continues to develop, we must ask ourselves how we, as the public, want to shape its role in society. We should not only improve AI models, but also make sure that AI is developed and used responsibly.
A number of companies are already taking action. They are judging the data, rules, and beliefs that shape the behaviour of AI models. Still, we cannot expect the companies to do all the work. As long as AI is trained on human data, it will reflect human behaviour. That means we have to think carefully about the digital footprints we leave in the world. I may value privacy, but if I give it up in a heartbeat to visit a website, the algorithms may make a very different judgement of what I really want and what is good for me. If I want meaningful human connections yet spend more time on social media and less time in the physical company of my friends, I am indirectly training AI models about the true nature of humanity.
As AI becomes more powerful, we need to take increasing care to inscribe(铭刻) our principles(原则) into the record of our actions rather than allowing the two to diverge. Recognizing this allows us to make better decisions, but only when we are prepared to look closely and take responsibility(责任) for what we see.
(
A. To suggest a solution.
B. To stress a difference.
C. To challenge a practice.
D. To support a viewpoint.
(
A. Improve.
B. Appear.
C. Separate.
D. Repeat.
(
A. Guarding one's privacy against AI models.
B. Being mindful of our feeds into AI models.
C. Training algorithms to favour the latest data.
D. Designing algorithms to deal with unfairness.
(
A. AI isn't the problem; we are
B. AI: A tool to reshape our society
C. More open algorithms for better AI
D. Building trust in human-AI relationships
People are talking a lot about artificial intelligence(AI), viewing it as a force that could reshape how society works. But there is something important missing from this discussion. It isn't enough to ask how it will change us. We also need to understand how we shape AI and what it can tell us about ourselves.
Every AI model we develop mirrors our rules and expresses our beliefs. A few years ago, while looking for new workers, a famous company gave up an AI-powered tool after finding it unfavourable to women. The AI was not designed to behave this way. Instead, it was influenced by historical data favouring men. Similarly, a recent study found that lending algorithms(算法) often offer less favourable terms to people of colour, worsening long-standing unfairness in the lending industry. In both cases, AI isn't creating new biases(偏见); it is mirroring the ones that are already present.
These reflections(反映) give us an important chance to take a close look at ourselves. By making these problems seen and more pressing, AI challenges us to recognize and address what causes algorithmic bias. As AI continues to develop, we must ask ourselves how we, as the public, want to shape its role in society. We should not only improve AI models, but also make sure that AI is developed and used responsibly.
A number of companies are already taking action. They are judging the data, rules, and beliefs that shape the behaviour of AI models. Still, we cannot expect the companies to do all the work. As long as AI is trained on human data, it will reflect human behaviour. That means we have to think carefully about the digital footprints we leave in the world. I may value privacy, but if I give it up in a heartbeat to visit a website, the algorithms may make a very different judgement of what I really want and what is good for me. If I want meaningful human connections yet spend more time on social media and less time in the physical company of my friends, I am indirectly training AI models about the true nature of humanity.
As AI becomes more powerful, we need to take increasing care to inscribe(铭刻) our principles(原则) into the record of our actions rather than allowing the two to diverge. Recognizing this allows us to make better decisions, but only when we are prepared to look closely and take responsibility(责任) for what we see.
(
D
) 1.Why does the writer introduce the two examples in Paragraph 2?A. To suggest a solution.
B. To stress a difference.
C. To challenge a practice.
D. To support a viewpoint.
(
C
) 2.What does the word "diverge" in the last paragraph most probably mean?A. Improve.
B. Appear.
C. Separate.
D. Repeat.
(
B
) 3.According to the passage, what is a good example of shaping AI responsibility?A. Guarding one's privacy against AI models.
B. Being mindful of our feeds into AI models.
C. Training algorithms to favour the latest data.
D. Designing algorithms to deal with unfairness.
(
A
) 4.Which of the following is the best title for this passage?A. AI isn't the problem; we are
B. AI: A tool to reshape our society
C. More open algorithms for better AI
D. Building trust in human-AI relationships
答案
1.D 2.C 3.B 4.A
解析
【分析】
这是一篇议论文类阅读理解，解题时首先梳理全文核心论证逻辑：文章跳出大众讨论AI如何改变人类的常规视角，提出核心观点——AI本质是人类自身规则、信念的映射，AI暴露的问题本质是人类自身的问题。之后逐个对应题目考点思考：第1题是议论文例证作用题，直接定位例子前后的论点即可判断；第2题是词义猜测题，结合划线词所在句的转折逻辑推导词义；第3题是细节理解题，定位到文中关于“负责任开发使用AI”的相关表述匹配选项；第4题是主旨大意题，排除只覆盖局部内容的干扰项，选出匹配全文核心论点的标题即可。
【解析】
1. 第1题：第二段开篇先提出核心观点“我们开发的每一个AI模型都映射人类的规则、表达人类的信念”，随后举了招聘AI歧视女性、借贷算法歧视有色人种两个实例，最终总结点明AI并没有创造新偏见，只是复刻了人类社会已有的偏见，两个例子的作用就是支撑前文提出的观点，因此选D。A（提出解决方案）、B（强调差异）、C（挑战惯例）均不符合例证的作用逻辑。
2. 第2题：划线词所在句含义为“随着AI变得越来越强大，我们需要谨慎地将自身的原则铭刻到行为记录中，而不是让原则和我们的行为二者______”，前文强调我们要让自身行为和秉持的原则保持统一，转折后的反向逻辑就是不让二者分离、产生分歧，因此diverge的含义最接近Separate（分离），选C。A（提升）、B（出现）、D（重复）均不符合语境逻辑。
3. 第3题：根据第四段内容“只要AI是基于人类数据训练的，它就会反映人类行为。这意味着我们必须仔细考虑自己在网络世界留下的数字足迹”，可知负责任地塑造AI的正确做法是留意我们输入给AI模型的各类行为数据，对应选项B。A（完全防范隐私不被AI获取）、C（训练算法优先使用最新数据）文中均未提及，D（设计算法解决不公平）只是局部优化手段，不属于文中强调的从人类源头出发的负责任做法。
4. 第4题：全文始终围绕核心论点展开：AI本身不会主动产生偏见和问题，它所有的偏差表现都是人类自身偏见、行为习惯的映射，问题的根源实际出在人类自身，因此最贴合主旨的标题是A。B仅对应文章开篇引入的次要内容，C、D的核心话题“开放算法”“人机信任”均未在文中重点论述，属于偏离主旨的干扰项。
【知识点】
议论文阅读 词义猜测 主旨大意理解
【点评】
本题属于典型的观点类议论文阅读，考点覆盖了阅读理解的常考题型，解题的关键是抓住议论文“论据服务于论点”的核心逻辑，不要被局部细节信息误导，要从全文的核心论述主线出发判断选项，能很好地考查学生的逻辑梳理和主旨提炼能力。
【难度系数】
0.6
这是一篇议论文类阅读理解，解题时首先梳理全文核心论证逻辑：文章跳出大众讨论AI如何改变人类的常规视角，提出核心观点——AI本质是人类自身规则、信念的映射，AI暴露的问题本质是人类自身的问题。之后逐个对应题目考点思考：第1题是议论文例证作用题，直接定位例子前后的论点即可判断；第2题是词义猜测题，结合划线词所在句的转折逻辑推导词义；第3题是细节理解题，定位到文中关于“负责任开发使用AI”的相关表述匹配选项；第4题是主旨大意题，排除只覆盖局部内容的干扰项，选出匹配全文核心论点的标题即可。
【解析】
1. 第1题：第二段开篇先提出核心观点“我们开发的每一个AI模型都映射人类的规则、表达人类的信念”，随后举了招聘AI歧视女性、借贷算法歧视有色人种两个实例，最终总结点明AI并没有创造新偏见，只是复刻了人类社会已有的偏见，两个例子的作用就是支撑前文提出的观点，因此选D。A（提出解决方案）、B（强调差异）、C（挑战惯例）均不符合例证的作用逻辑。
2. 第2题：划线词所在句含义为“随着AI变得越来越强大，我们需要谨慎地将自身的原则铭刻到行为记录中，而不是让原则和我们的行为二者______”，前文强调我们要让自身行为和秉持的原则保持统一，转折后的反向逻辑就是不让二者分离、产生分歧，因此diverge的含义最接近Separate（分离），选C。A（提升）、B（出现）、D（重复）均不符合语境逻辑。
3. 第3题：根据第四段内容“只要AI是基于人类数据训练的，它就会反映人类行为。这意味着我们必须仔细考虑自己在网络世界留下的数字足迹”，可知负责任地塑造AI的正确做法是留意我们输入给AI模型的各类行为数据，对应选项B。A（完全防范隐私不被AI获取）、C（训练算法优先使用最新数据）文中均未提及，D（设计算法解决不公平）只是局部优化手段，不属于文中强调的从人类源头出发的负责任做法。
4. 第4题：全文始终围绕核心论点展开：AI本身不会主动产生偏见和问题，它所有的偏差表现都是人类自身偏见、行为习惯的映射，问题的根源实际出在人类自身，因此最贴合主旨的标题是A。B仅对应文章开篇引入的次要内容，C、D的核心话题“开放算法”“人机信任”均未在文中重点论述，属于偏离主旨的干扰项。
【知识点】
议论文阅读 词义猜测 主旨大意理解
【点评】
本题属于典型的观点类议论文阅读，考点覆盖了阅读理解的常考题型，解题的关键是抓住议论文“论据服务于论点”的核心逻辑，不要被局部细节信息误导，要从全文的核心论述主线出发判断选项，能很好地考查学生的逻辑梳理和主旨提炼能力。
【难度系数】
0.6
