The AI model Deepseek was all over the news in January 2025. Deepseek's developers claimed its costs to develop and run were a fraction of the costs for large popular AI models, which caused a big disruption in the stock market. But how true are these claims? In the first of a three-part series on AI, host Michele Bousquet dives in and give Deepseek a spin, and debunks some of the hype around this new export from China.

Resources:

• DeepSeek-V3 Technical Report
• The Extreme Cost Of Training AI Models
• How Much Did It Cost to Train GPT-4? Let’s Break It Down
• OpenAI says it has proof DeepSeek used its technology to develop its AI model
• NASDAQ:NVDA Stock Falls on DeepSeek Fears – Is Wall Street Missing a 71% Upside?

Send us a text

Support the show

Join our Patreon to listen ad-free!