DeepSeek and the US-China AI Race

Modern War Institute

John Amble

Government, News

4.7 • 798 Ratings

🗓️ 9 February 2025

⏱️ 49 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

When the Chinese company DeepSeek recently released an artificial intelligence model called R1, its surprisingly advanced capability and the efficiency with which DeepSeek claimed to have trained the model sparked a wave of discussion about how rapidly Chinese AI development was progressing. But how is the competition to achieve superior AI technologies between China and the United States shaping up? And, given that the governments of both countries envision important military applications of AI, what implications does this have for the broader US-China strategic competition? To explore these questions and more, John Amble is joined by Bill Hannas, lead analyst at Georgetown University’s Center for Security and Emerging Technology (CSET), and Sam Bresnick, a CSET research fellow. The MWI Podcast is produced with the generous support of the West Point Class of 1974.

Transcript

Click on a timestamp to play from that location

0:00.0	But clearly there's a race toward artificial general intelligence, AGI, and it's recognized
0:09.6	as such by top players on both sides.
0:13.7	I think there's a real chance of automation bias within the Chinese military, that there's
0:19.2	concern about making decisions, there's concern about,
0:22.8	you know, going against the technology. Denying chips and other technology to China, at best,
0:30.7	it delays China's progress and worst, and I think that's what we're seeing here, it incentivizes
0:35.9	Chinese scientists to develop a genius workarounds.
0:40.6	Hey, welcome back to the Modern War Institute podcast.
0:43.6	Brought to you with the generous support of the West Point class of 1974.
0:47.6	I'm John Amble.
0:48.6	And on this episode, we're digging into the artificial intelligence competition between the United States and China.
0:55.8	When the Chinese company DeepSeek recently released an AI model called R1, experts quickly noted
1:01.3	that its capability is largely on par with some of the most advanced large language
1:05.0	models developed by leading US companies like OpenAI.
1:08.9	This and the fact that DeepSeek claims to have trained its model
1:11.3	much more cheaply than U.S. companies sparked a wave of discussion about how rapidly Chinese
1:16.2	AI development was progressing. And because both China and the United States see a role for military
1:22.1	use of AI, it's a really important discussion. To explore that subject, I'm joined on this episode by two experts from Georgetown University's
1:30.4	Center for Security and Emerging Technology.
1:33.1	Bill Hannes is lead analyst at CSET, and Sam Bresnick is a CSET research fellow.
1:38.4	Before we get to the discussion, a couple of quick notes.
1:41.1	First, a very special thank you to the West Point class of 1974 for their
	...

Please login to see the full transcript.

Previous episode | Next episode

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

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