📼 China vs. the U.S.—Who Will Win the AI Race? Lex Fridman's New Podcast
How AI happens to be at the center of the "new Cold War" and how the world would change with the emergence of strong AI (AGI)—these are among the issues Lex Fridman discusses with Dylan Patel, founder of research company SemiAnalysis, and Nathan Lambert of The Allen Institute for AI, in his podcast's new episode.
Highlights:
🛑 Chinese startup DeepSeek has been able to leverage resources even with the U.S. now restricting chip exports, Patel says. Nathan Lambert believes current U.S. sanctions are unlikely to prevent China from creating new AI models—but they make it difficult for the industry to scale.
"People in AI have been worried that this is going towards a Cold War or already is," Nathan Lambert says.
🛑 China's industrial capabilities far surpass those of the U.S., Patel says. The huge data centers the U.S. wants to build as part of Stargate will consume gigawatts of power, but that would still not exceed the giant Chinese industrial facilities.
"China, if they wanted to build the largest data center in the world, if they had access to the chips, could. So it's just a question of when, not if," Dylan Patel warns.
🛑 Today, the default AI race leaders are Google, with the best infrastructure, and OpenAI, with the best AI models for people to use, experts agree.
🛑 Strong AI will fundamentally change society, Lambert believes. The OpenAI o3 model has a spectacular coding performance that helps AI companies to experiment better.
🛑 Still, Patel insists the economic revolution won't happen in a snap of a finger. Implementing super-powered AI will require time, resources, and relevant infrastructure.
🛑 AGI might already be there in part, Patel notes. For example, in recent elections in India and Pakistan, voters got AI voice calls and thought they were talking to the politicians.
🛑 At some point, every powerful human will want a brain-computer interface to interact with the AGI, Dylan Patel predicts. Those human-machine amalgamations enable an individual human to have more impact on the world, and that impact can be both positive and negative.
📱 You can listen to the full interview here.
More on the topic:
➡️ The Success of DeepSeek: How China's Open Source Model Challenges ChatGPT
