[FCE] The Big Idea: why we should embrace AI doctors | Books | The Guardian

收听本期播客

阅读正文

Artificial Intelligence (AI) is increasingly being seen as a potential solution to some of the biggest challenges in healthcare. Doctors, despite their expertise, are human and face significant pressures that can lead to errors. In England, for example, around 5% of visits to general practitioners result in incorrect diagnoses, putting countless patients at risk. In the United States, nearly 800,000 individuals suffer serious injury or death each year due to such mistakes. With these alarming figures, many are asking whether AI could help reduce these errors and improve patient outcomes.

One of the main issues in healthcare is the sheer volume of information doctors must handle. Medical knowledge is expanding so rapidly that much of what students learn during training becomes outdated by the time they graduate. Additionally, factors like exhaustion and stress can affect a doctor’s performance. Studies have shown that as the workday progresses, doctors are more likely to make mistakes, such as prescribing unnecessary medication or overlooking critical tests. Access to healthcare is another concern, as many people—particularly those who are most vulnerable—struggle with long waiting times or difficulties in reaching medical facilities.

AI offers a promising alternative in addressing these problems. Unlike humans, AI systems can analyse enormous amounts of data without ever losing focus. They can detect patterns and identify conditions, especially rare ones, that might be missed by a tired or overworked doctor. In a 2023 study, an AI tool named ChatGPT-4 was tested on 50 medical cases. It correctly diagnosed all common cases by its second suggestion and identified 90% of rare conditions by its eighth attempt, often surpassing human doctors. Real-life cases highlight this potential too. A young boy, after visiting 17 doctors over three years with chronic pain and receiving no clear diagnosis, had his condition identified by ChatGPT as a rare disorder—a diagnosis later confirmed by specialists.

Beyond diagnosis, AI could also improve access to healthcare. Virtual consultations through apps, such as those planned for the UK’s National Health Service, could allow millions to receive medical advice without waiting for appointments. However, challenges remain. Not everyone has access to the internet or feels comfortable using digital tools, meaning that gaps in technology skills need to be addressed. While AI is not flawless and concerns about errors or bias persist, it presents an opportunity to enhance a healthcare system that often struggles to meet patients’ needs.

阅读练习

1. What is the main purpose of the article?

  • A. To explain how doctors are trained
  • B. To discuss the potential of AI in solving healthcare issues
  • C. To describe the history of medical errors
  • D. To argue against the use of technology in medicine

2. According to the article, what is one reason doctors make mistakes?

  • A. They do not have enough patients
  • B. They are often tired or stressed
  • C. They refuse to use new technology
  • D. They lack proper qualifications

3. What does the article say about the performance of ChatGPT-4 in a 2023 study?

  • A. It failed to diagnose any rare conditions
  • B. It was less accurate than most doctors
  • C. It identified most rare conditions by its eighth attempt
  • D. It only worked on common cases after several tries

4. How could AI improve access to healthcare, according to the text?

  • A. By training more doctors to use technology
  • B. By offering virtual consultations through apps
  • C. By reducing the cost of medical equipment
  • D. By building more hospitals in rural areas

5. What is one challenge mentioned in using AI for healthcare?

  • A. AI systems are too expensive to develop
  • B. Not everyone has the skills to use digital tools
  • C. AI cannot process large amounts of data
  • D. Doctors are unwilling to work with AI