ChatGPT, the bot driven by artificial intelligence, recently passed three U.S. medical exams.

Designed by research lab OpenAI, it’s mulled the fundamentals of business and the law and expounded on them in essays.

Its fluidity in discourse and its broad knowledge has some, including journalists, nervous about the technology’s capacity to replace humans in the workforce.

Ruth got up early to catch ChatGPT before high demand for its oracle-like powers created a queue, and she was able to get its perspective on the future of AI in medicine.

The interview has been edited for length.

Can AI replace doctors?

No, AI cannot replace doctors. While AI has the potential to improve many aspects of medical care, it does not have the capability to replace human doctors. AI can assist doctors in diagnosing diseases, making predictions and providing recommendations, but it does not have the capacity for human empathy, emotional intelligence and ethical decision-making that are essential for the practice of medicine.

Additionally, AI algorithms are only as good as the data they are trained on, and human judgment is still necessary to interpret and validate the results.

Is it true you can pass medical licensing exams?

It is possible that an AI language model like ChatGPT, which has been trained on vast amounts of text data, including medical information, could perform well on some multiple-choice questions found on medical licensing exams.

However, AI is not capable of replacing human doctors or passing a comprehensive medical licensing exam.

A medical licensing exam assesses not just medical knowledge but also clinical skills and the ability to diagnose and treat patients in a safe and effective manner. These abilities require human judgment, emotional intelligence and critical thinking, which AI algorithms do not possess.

Can you tell me about all the ways AI is being used in medicine?

Artificial intelligence (AI) is being used in various ways in medicine. Some of the most common uses are:

1. Diagnostic imaging: AI algorithms can help analyze medical images such as CT scans, MRI and X-rays to identify anomalies or assist in making a diagnosis.
2. Electronic health records: AI algorithms can help extract information from EHRs to assist in patient care, population health management and medical research.
3. Drug discovery and development: AI can help identify new drug targets, predict drug interactions and side effects and optimize clinical trial design.

These are just a few examples of how AI is being used in medicine, and the field is rapidly evolving, with new applications being developed all the time.

Is it possible that artificial intelligence could also increase medical errors?

Yes, it is possible. There are several potential sources of error, such as:

1. Bias in training data. If the training data contains biases or inaccuracies, the AI system could make incorrect predictions or recommendations.
2. Limited context. AI algorithms can process vast amounts of data, but they may not have access to all of the relevant information about a patient. This could lead to incorrect predictions or recommendations if the AI system is missing important context.
3. Lack of transparency. AI algorithms can be complex and difficult to understand, making it difficult to determine how they arrived at a particular conclusion.

Ongoing monitoring and evaluation of AI systems are necessary to ensure that they are functioning as intended and to minimize the potential for medical error.

This is where we explore the ideas and innovators shaping health care.

Robots with locust antennae may replace bomb-sniffing dogs at airports, if a recent invention by Israeli scientists is widely adopted. Which do you want checking your luggage?

Share any other thoughts, news, tips and feedback with Ben Leonard at [email protected], Ruth Reader at [email protected], Carmen Paun at [email protected] or Erin Schumaker at [email protected].

Send tips securely through SecureDrop, Signal, Telegram or WhatsApp.

Today on our Pulse Check podcast, Adam Cancryn tells Ben about his exit interview with David Kessler, who reflected on nearly three years of fighting a pandemic as President Joe Biden’s chief science officer — and the political fallout. Plus, Rep. David Schweikert (R-Ariz.), co-chair of the Telehealth Caucus, on how technology can become a bigger part of health care delivery.

While lead exposure has dramatically decreased in high-income countries since the 1970s because of the phaseout of leaded gasoline, it’s still highly lethal in the developing world.

The death toll from lead was about 50 percent greater than from malaria in 2019, according to the Seattle-based Institute for Health Metrics and Evaluation.

In addition, the chemical affects children’s neurological development, leading to lifelong cognitive deficits and attention and behavioral challenges.

Up to half the children in developing countries are exposed to lead at levels so high that the World Health Organization recommends a public health intervention.

A new report from the Center for Global Development, a think tank focused on reducing poverty, says the causes of lead poisoning include the recycling of batteries containing lead acid, mining and ore smelting, lead paint, cookware, cosmetics, toys, consumer goods and contaminated spices.

“The effects of lead poisoning on cognition and behavior, combined with its high prevalence, suggest that lead exposure is likely to have a substantial impact on overall educational attainment, crime, violence, and potentially economic growth,” wrote authors Rachel Silverman Bonnifield, a senior fellow at the center, and Rory Todd, a research assistant.

They said the countries most affected should:

— Measure and monitor lead-exposure levels

— Assess who’s exposed to identify the causes

— Strengthen health systems so they have the ability to diagnose and treat lead poisoning

— Improve nutrition to limit lead absorption

The world’s richest countries, including the United States, also have a role to play, the report notes.

Wealthy countries can help by:

— Restating the importance of reducing lead exposure

— Adopting standards through an international agreement

—Expanding financial support for poorer countries in identifying the local prevalence of lead exposure and its sources

— Complying with international conventions on the movement of waste containing lead across borders

Improving data-sharing is often touted as a way to better care for patients.

A study published in the Annals of Family Medicine supports that view.

Researchers analyzed the 2021 health records of 8,000 discharged hospital patients who visited their primary care doctor within 30 days of release. The data was taken from a single health information exchange in El Paso County, Texas.

Striking results: When primary care doctors looked up patients’ records in the exchange two days before or after the visit, emergency department readmissions dropped 53 percent and rehospitalizations dropped 61 percent compared to when they didn’t.

“When used, the information can be remarkably beneficial to increasing the time to future hospital use,” the researchers wrote, presumably because doctors were better able to tailor care.

Even so: The study also found low adoption: Doctors looked up in the exchange only about 3 percent of patients who had a primary care visit within 30 days of discharge.

This low usage rate could be because doctors didn't know patients were hospitalized and their patient didn't tell them. At the same time, the doctor might not think to ask.

The exchange in El Paso is discussing setting up an alert system to notify primary care doctors when patients are admitted to hospitals, the study said.