AI: The Peril And Opportunity Of Artificial Superintelligence : 1A : NPR

A smartphone displaying the OpenAI logo above the word “Anthropic” is seen beside a lit headset

Imen Ben Youssef / Hans Lucas/Getty Images

hide caption

toggle caption

Imen Ben Youssef / Hans Lucas/Getty Images

Artificial Intelligence is advancing at a dizzying pace. One analysis shows it doubling its abilities every seven months.

And it’s surpassed humans in more than just trivia and Chess. Last year, an AI from Microsoft solved complex medical cases with 85% accuracy, far about the 20% average for experienced physicians. And a recent Stanford report found that some of the newest A-I systems now match or beat the average human expert on PhD-level science questions.

But what happens when A-I is better and smarter than the brightest among us at every task? That’s called superintelligence.

Researchers disagree about how close we are to that sci-fi goal: is it years, or decades—or possible at all? And what happens if that genie-in-a-bottle is let loose? Some say the risk is as existential as total human extinction.

We’ll discuss the biggest promise – and peril – of AI’s advancement beyond humans.

Find more of our programs online. Listen to 1A sponsor-free by signing up for 1A+ at plus.npr.org/the1a.

Source