Can we always trust the results we get from AI? Researchers from the University of Cambridge and the University of California Santa Barbara say no. They have shown that there are some problems that even the most powerful AI cannot reliably solve, no matter how much data it is given.

Many real-world systems, like those in oceans, the human brain, or robots, are too complex to describe neatly with equations (方程式). So researchers often use machine learning to study how they behave. But these AI methods do not always work well. Sometimes they return unreliable results or poor predictions. Sometimes, however, providing reliable solutions may be fundamentally impossible for some problems, even with infinite data.

The researchers designed special systems to test AI. These systems were built to find out exactly where and why AI prediction breaks down. They identified two main reasons why machine learning fails on complex systems. Either the algorithm cannot tell when it has seen enough data, or patterns in the system are hidden and hard to distinguish.

Their results may also help explain why AI chatbots can be accurate in the short term but drift or produce false information over time. The researchers found that chaotic systems are especially problematic. When a system is chaotic, meaning tiny differences in starting conditions lead to very different results, short-term prediction can be accurate, but long-term prediction becomes unreliable. Small changes in a question can send a chatbot down a completely different path. The answer looks reasonable word by word but produces unrealistic information over longer outputs.

1.1. What did the researchers find about AI?

A It can handle every problem.

B It fails on certain problems.

C It always gives wrong answers.

D It needs more training data.

解析：选B。B细节理解题。根据第一段原文“there are some problems that even the most powerful AI cannot reliably solve”可知研究者的发现是：有些问题即使是再强大的AI也无法可靠解决。A颠倒是非（与原文结论相反），C以偏概全（原文说“not always work well“，而非”always wrong”），D无中生有（原文未说需要更多训练数据）。故选B。

2.2. Why does machine learning break down on complex systems?

A Hidden patterns and insufficient data awareness.

B Weak computers and poor models.

C It always gives wrong answers.

D It needs more training data.

解析：选A。A推理判断题。根据第三段“They identified two main reasons why machine learning fails on complex systems. Either the algorithm cannot tell when it has seen enough data, or patterns in the system are hidden and hard to distinguish.”可知两大原因：①算法无法判断数据是否足够（insufficient data awareness）；②系统模式隐藏难以分辨（hidden patterns）。B无中生有（未提及电脑性能），C张冠李戴（方程式是传统方法的局限，非机器学习失败的原因），D无中生有（未提及目标不明确或团队未经训练）。故选A。

3.3. The underlined word “chaotic” in Paragraph 4 probably means ______.

A extremely disordered

B perfectly organized

C exactly repeated

D slowly changing

解析：选A。A词义猜测题。根据第四段原文“tiny differences in starting conditions lead to very different results”（起始条件的微小差异导致截然不同的结果）可推断chaotic意为“混乱的、无序的”。B颠倒是非（organized与chaotic相反），C颠倒是非（chaotic系统不会精确重复），D无中生有（未提及变化速度）。故选A。

4.4. What can we infer about the new algorithm?

A Very expensive to run.

B Hard to operate well.

C Quite efficient to use.

D Ignoring real-world data.

解析：选C。C推理判断题。根据最后一段“outperformed current leading AI models at a fraction of the cost, running on a standard laptop”可知新算法成本低、性能强、普通电脑即可运行，可推断其效率高。A颠倒是非（原文“at a fraction of the cost”说明成本低），B过度推理（未提及操作难度），D颠倒是非（算法恰恰基于真实海冰数据）。故选C。