D 热考内容 时事新闻
KEY→P18
范畴:人与社会 主题:科学技术 体裁:说明文 难度:★★★★ 建议用时:7分钟 评价:/4
Do you know about the AI chatbot ChatGPT? It is getting smarter now! On August 7, 2025, US company OpenAI introduced GPT-5, calling it their "smartest, fastest, most useful model yet".
OpenAI said that the new AI is cleverer than older ones. "GPT-5 is the first time that it really feels like talking to an expert in any topic, like a PhD-level expert," Sam Atman, OpenAI's CEO, told the media.
How does it get so smart? According to OpenAI, GPT-5 learns using neural networks(神经网络). This permits the chatbot to look for patterns in mountains of text, code and pictures. The AI
also "thinks" differently. It chooses between quick answers for simple questions, like math facts, and a slower, deeper thinking way for harder problems.
OpenAI said that GPT-5 is better at getting the facts straight, too. During normal chats, GPT-5 makes about 45 percent fewer mistakes than GPT-4o. When it is in the mode of deep thinking, its mistakes drop by about 80 percent compared with the older o3 model.
Another cool feature is that GPT-5 has different "personalities(人格)", like Cynic, Robot, Listener or Nerd. People can choose among them for different tasks.
Still, GPT-5 isn't perfect. Google's Gemini is better at giving the latest news, and China's DeepSeek does a good job at much lower costs.
()1. Why did Sam Altman compare GPT-5 to an expert?
A. To show GPT-5 is smarter than experts.
B. To talk about what GPT-5 does well in.
C. To show the improvement of ChatGPT.
D. To introduce why the AI chatbot is smart.
()2. How does GPT-5 work according to Paragraph 3?
A. By learning through neural networks.
B. By using other networks to help think.
C. By looking for answers with patterns.
D. By using pictures to design its patterns.
()3. What can we learn about GPT-5 from Paragraph 4?
A. It never makes mistakes while thinking.
B. It gives more correct answers than others.
C. It can check facts before giving answers.
D. It can get facts faster than the old models.
()4. What might the writer of the passage agree with?
A. GPT-5 is better than China's DeepSeek at deep thinking.
B. GPT-5 does better in getting the latest news than Gemini.
C. GPT-5 is not strong enough to deal with all our problems.
D. GPT-5 with different personalities has much lower costs.

本文主要介绍了OpenAI公司推出的GPT-5智能聊天机器人。
1. C 推理判断题。根据第二段第一句“OpenAI said that the new AI is cleverer than older ones.”可知,OpenAI说新的AI比旧的更聪明,所以此处Sam Altman把GPT-5比作专家是为了展示ChatGPT的进步。
2. A 细节理解题。根据第三段中的“According to OpenAI, GPT-5 learns using neural networks.”可知,GPT-5通过使用神经网络来学习。
3. B 推理判断题。根据第四段中的“When it is in the mode of deep thinking, its mistakes drop by about 80 percent compared with the older o3 model.”可知,当它在进行深度思考时,其错误率与旧型号相比下降了约80%,说明它给出的答案比其他的更准确。
4. C 推理判断题。根据最后一段内容可知,GPT-5仍然不完美,谷歌的Gemini在提供最新消息方面做得更好,而中国的DeepSeek则以更低的成本做得很好,因此作者认为GPT-5还没有足够强大到能解决所有的问题。