Reading the GPT Leaves

Politix

Politics, News Commentary, News

4.6 • 1.4K Ratings

🗓️ 16 December 2022

⏱️ 63 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

If you could go back in time and had the power to stop the social media revolution in its tracks, what would you do? Would you let things proceed exactly as they did? Or would you try to warn people? We’re at a similar fork in the road right now with artificial intelligence, and the recent advent of OpenAI’s ChatGPT. It’s fun to play around with, but when you do, it becomes clear that the horizons have changed. And if AI like ChatGPT becomes as ubiquitous as smartphones and earbuds and the Internet of Things, it’s going to change the world. A lot. How can we impose order or mitigate the risks around artificial intelligence? Can these early stage innovations generally do more harm than good in the near-term—that is, before the machines get smart enough to build weapons and enslave us? New York Times technology reporter and host of the Hard Fork podcast Kevin Roose joins Brian Beutler to talk about what upheavals AI might bring, including the amazing ones, the scary ones, and the robot apocalypse ones.

Transcript

Click on a timestamp to play from that location

0:00.0	Hi, everyone. Welcome to Positively Dreadful with me, your host, Brian Boyler. So, here's
0:23.6	a question I'd like you all to ponder for about five seconds wherever you happen to be listening.
0:28.4	If you could go back in time to the days just before Friendster and then Myspace and then Facebook
0:33.1	snake their way through the world and you had the power to do anything from just stop it in its
0:38.8	tracks. So, the social media revolution would never happen. To nothing at all, what would you do?
0:44.5	Would you hit the kill switch and spare us all from the social web? Would you let things proceed
0:50.5	exactly as they did? Would you try to warn people? Look, this is going to start out really cool and
0:56.5	chill, but soon it will consume way more of our lives and societies than is probably healthy. It'll
1:02.9	be a breeding ground for hatreds and anxieties. We should probably get ahead of that now. I don't
1:08.9	think of myself as any kind of bloodite, but I'm sure I would intervene somehow, maybe even by
1:14.9	shutting the whole thing down. And I've been thinking about this a lot in the past couple weeks, not just
1:20.5	because of Elon Musk or every other bad thing that's come out of social media in the past several
1:26.3	years, but because I think we may be at one of these fork-and-the-road moments right now. And I
1:32.2	think there's hope that we can be a bit less naive about what risks we might be courting now
1:38.0	than we were in the innocent days of the early web. So, like a lot of people, I've been playing around
1:44.4	with chat GPT. Chat GPT refers specifically to the generative pre-trained transformer,
1:50.8	but if you've heard about it, you probably know it as like the new AI chatbot. It's a tool built
1:57.4	by the company OpenAI, which builds artificial intelligence, and chat GPT is basically just as I
2:04.1	described. It's a web-based chat window, but instead of talking to a person, you're talking to a
2:09.5	machine. And the machine is built by humans and refined by humans, but most of the learning chat GPT
2:16.3	has done to inform its answers and write them in proper English and to create a remarkable degree
2:23.6	of very similar to all of that is automated. Basically, if it was on the web in the year 2021,
	...

Please login to see the full transcript.

Previous episode | Next episode

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

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