At a meeting, two people may use the same AI tool and leave with very different minds. One asks it for a quick answer, copies the result, an...

At a meeting, two people may use the same AI tool and leave with very different minds. One asks it for a quick answer, copies the result, and moves on. The other treats it less like a shortcut and more like a demanding partner, testing, revising, and arguing with it until the work begins to deepen. From the outside, both seem productive. From the inside, however, one is merely finishing a task, while the other is learning how to think with greater range and care.

That difference sits near the heart of human AI collaboration. MIT Sloan reports that researchers studying consultants at Boston Consulting Group identified three broad ways people work with generative AI: cyborgs, who collaborate closely with it, centaurs, who guide it in a more controlled way, and self automators, who hand over most of the task. The labels matter because they show that the real issue is not use alone, but manner of use. A tool may extend judgment, or it may quietly replace it.

The most interesting finding is not that AI helps, since that is hardly a surprise now. It is that different habits produce different kinds of growth. MIT Sloan says centaurs achieved the highest accuracy and strengthened their domain expertise through focused questions, while cyborgs became better at solving problems with AI itself. Self automators, by contrast, saw no gains in skill. Not through failure, but through effortless polish, can dependence become dangerous. Work may look smoother even as understanding grows thinner.

So the future of collaboration may depend less on whether humans keep control in some dramatic, cinematic sense, and more on whether they remain mentally present. MIT Sloan also notes that organizations should actively shape training and workflows so employees learn when to automate and when not to. That is a practical lesson, but also a human one. A good partnership with AI should lighten routine burdens without hollowing out judgment. Otherwise, the machine does not steal our work in one stroke. We slowly stop showing up to it ourselves.

[Adapted from MIT Sloan]

Question 23. Which of the following is NOT mentioned in paragraph 1 as a way people may work with the same AI tool?

A. One user copies the answer quickly and moves on.

B. One user argues with the tool to refine the work further.

C. Both users appear productive to outside observers.

D. Both users are required to follow the same method of revision.

Question 24. In paragraph 1, the word “deepen” is closest in meaning to __________.

A. improve        B. expand        C. continue        D. simplify

Question 25. The word “it” in paragraph 2 refers to __________.

A. generative AI        B. judgment        C. manner of use        D. the real issue

Question 26. In paragraph 4, the phrase “hollowing out” is OPPOSITE in meaning to __________.

A. remaking        B. strengthening        C. refreshing        D. changing

Question 27. Which of the following best paraphrases the sentence in paragraph 3?

A. AI-supported work can seem more efficient even while the user’s real understanding becomes weaker.

B. Better-looking work usually proves that users have developed stronger knowledge and better judgment.

C. When tasks become easier to complete, people naturally gain a deeper grasp of the work involved.

D. Smooth performance with AI mainly shows that users have learned to solve problems more independently.

Question 28. Which of the following is best supported by paragraph 3 about self-automators?

A. Their reliance on AI can make the work appear stronger than their actual understanding.

B. Their main weakness is that they refuse to let AI help with demanding tasks.

C. They gain the greatest benefit from AI because they save the most time.

D. They are more likely than other users to deepen expertise through careful prompting.

Question 29. Which paragraph focuses on different styles of working with generative AI and why those styles matter more than simple use?

A. Paragraph 1        B. Paragraph 2        C. Paragraph 3        D. Paragraph 4

Question 30. Which paragraph emphasizes that the future quality of human–AI collaboration depends on people staying mentally engaged rather than merely keeping formal control?

A. Paragraph 1        B. Paragraph 2        C. Paragraph 3        D. Paragraph 4