When ChatGPT Broke an Entire Field

The Quanta Podcast

Quanta Magazine

Life Sciences, Science, Physics

4.7 • 638 Ratings

🗓️ 29 July 2025

⏱️ 30 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

The study of natural language processing, or NLP, dates back to the 1940s. It gave Stephen Hawking a voice, Siri a brain and social media companies another way to target us with ads. In less than five years, large language models broke NLP and made it anew.

In 2019, Quanta reported on a then-groundbreaking NLP system called BERT without once using the phrase “large language model.” A mere five and a half years later, LLMs are everywhere, igniting discovery, disruption and debate in whatever scientific community they touch. But the one they touched first — for better, worse and everything in between — was natural language processing. What did that impact feel like to the people experiencing it firsthand?
Recently, John Pavlus interviewed 19 current and former NLP researchers to tell that story. In this episode, Pavlus speaks with host and Quanta editor in chief Samir Patel about this oral history of “When ChatGPT Broke an Entire Field.”

Each week on 𝘛𝘩𝘦 𝘘𝘶𝘢𝘯𝘵𝘢 𝘗𝘰𝘥𝘤𝘢𝘴𝘵, 𝘘𝘶𝘢𝘯𝘵𝘢 𝘔𝘢𝘨𝘢𝘻𝘪𝘯𝘦 editor in chief Samir Patel speaks with the people behind the award-winning publication to navigate through some of the most important and mind-expanding questions in science and math.

Audio coda from LingoJam

Transcript

Click on a timestamp to play from that location

0:00.0	Before ChatGPT or Claude or Alexa, there was Eliza.
0:10.8	Eliza was a computer program written by MIT computer scientist Joseph Weisenbaum around,
0:17.6	wait for it, 1966.
0:20.5	It was a kind of rules-based chatbot that recognized keywords and phrases and had
0:26.6	programmatic responses designed to keep a conversation moving.
0:30.9	It was an experiment.
0:32.6	But when it was fed a script, specifically to act like a therapist So, you know, it said things like,
0:38.5	how did that make you feel or tell me about your mother? People felt like it had feelings.
0:45.1	They made a connection with Eliza. And even today, this kind of emotional connection with a computer
0:50.6	is known as the Eliza Effect. Eliza was an early product of the field known as natural language processing,
0:59.2	or the area of computer science that bridges the way that people interact with the way
1:04.5	that computers interact.
1:06.5	And if there was a single field that was completely turned on its head, exploded, when modern large language models appeared, it was natural language processing. So what happens to the people in a field like that when it gets hit by an asteroid called ChatGPT.
1:37.6	Welcome to the Quanta podcast where we explore the frontiers of fundamental science and math.
1:40.7	I'm Samir Patel, editor-in-chief of Quantum Magazine.
1:46.3	Not long ago, we explored the impact of AI on the pursuits of science and math with a special issue we called science, promise, and peril in the age of AI. One of the centerpieces
1:52.8	of that package was a feature by Quanta contributing writer and science journalist John Pavlis
1:59.2	about how ChatGPT broke the world of natural
2:04.1	language processing, and he did it in the form of an urgent kind of emotional oral history.
2:10.6	John's here with us today to talk about the story and what he learned. Welcome to the show, John.
2:15.8	Hi, thanks. I'm really glad to be here.
2:19.2	So what's the big idea we're going to be exploring today? So this piece has a lot of big ideas in it, because
	...

Please login to see the full transcript.

Previous episode | Next episode

Disclaimer: The podcast and artwork embedded on this page are from Quanta Magazine, and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of Quanta Magazine and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.