Logo

The AI Model That Tanked the Stock Market

en

January 28, 2025

TLDR: On Monday, DeepSeek, an open source AI model developed in China, tanked the stock market and erased $589b from Nvidia's market cap. Zvi Mowshowitz, AI expert, discussed the potential threat to American AI giants like OpenAI and Anthropic, impact on export restrictions on US chips.

1Ask AI

In this special emergency episode of the Odd Lots podcast, hosts Joe Wiesenthal and Tracy Alloway dive deep into the dramatic sell-off in the stock market triggered by the emergence of DeepSeek, an open-source AI model from China. This model has caused significant market disruptions, particularly affecting tech giants like Nvidia, which saw a staggering $589 billion wiped from its market capitalization. In a discussion featuring AI expert Zvi Mowshowitz, the episode explores the implications of DeepSeek and its threat to American AI technology.

Key Points Discussed

The Impact of DeepSeek

  • Market Reaction: The stock market crash followed the announcement of DeepSeek, reducing Nvidia’s stock by 17% and raising concerns across the tech sector.
  • Cost Efficiency: DeepSeek was reportedly trained for a mere $5.5 million, utilizing optimized methods to achieve performance comparable to other AI models developed with far more financial resources.

Understanding DeepSeek

  • Open Source Philosophy: DeepSeek’s development arose from an ideological standpoint advocating for open access AI that serves the collective good, unlike many competitors which have moved to closed-source models.
  • Technological Innovations: The creators of DeepSeek employed aggressive optimization techniques that allowed them to maximize the output from limited resources, making it a formidable competitor in the AI landscape.

The Big Picture on AI Competition

  • Concerns for U.S. Tech Giants: The episode highlights how companies like OpenAI, Anthropic, and Meta could face serious competition due to DeepSeek’s lower overhead and open-source nature. It poses a significant threat not just because of cost but also due to potential shifts in AI utilization strategies.
  • AGI Aspirations: DeepSeek is driven by the vision of creating artificial general intelligence (AGI), which raises ethical and operational concerns regarding the accessibility of such powerful technology globally.

Insights from Zvi Mowshowitz

Efficiency vs. Accessibility

  • Zvi notes that while DeepSeek’s training cost appears low, the overall investment into its infrastructure and talent was substantial, indicating that the technology gap may not be as wide as it seems.
  • DeepSeek’s ability to operate effectively on a diverse range of GPUs positions it advantageously over Nvidia, which is heavily invested in specialized chips. This opens pathways for other companies to utilize AI more broadly without being restricted to high-cost hardware.

Future of AI Compute

  • Jevons Paradox: Zvi discusses the phenomenon where increased efficiency in resources (like compute power) leads to greater overall demand for those resources. This suggests that cheaper training methods could lead to an explosion in AI usage rather than a reduction in demand for powerful chips.
  • The podcast also touches on how custom chips versus general-purpose GPUs will shape the future of AI deployment, indicating that while Nvidia currently enjoys a dominant position, continuous innovation could disrupt this status.

Conclusion

The introduction of DeepSeek and its subsequent market effects highlight the fast-evolving landscape of AI technology and its market implications. Companies need to rethink their strategies as open-source models like DeepSeek change the rules of competition. As Zvi Mowshowitz emphasizes, understanding the long-term ramifications of such technological advancements is crucial for both investors and companies involved in AI development.

Takeaways for Readers

  • Stay informed about emerging AI technologies, especially open-source models, as they can influence market dynamics profoundly.
  • Consider the economic implications of AI advancements, including how efficiency gains may heighten demand rather than reduce it.
  • Follow ongoing debates around ethics, accessibility, and competition in the AI sector, as these will shape the future landscape of technology.

Was this summary helpful?

Recent Episodes

Marko Kolanovic Is Back With a Warning for Stocks

Marko Kolanovic Is Back With a Warning for Stocks

Odd Lots

Marko Kolanovic was one of the biggest names in markets, earning the nickname 'Gandalf' for some eerily prescient calls over the years. But last July he left his role as JPMorgan's chief global market strategist and co-head of global research, after missing out on a pretty epic rally. Since then, stocks have climbed higher with valuations increasingly stretched. So what does Marko think of the market now? In this episode we talk about his outlook the market, the impact of AI including the new DeepSeek model out of China, plus his own research and analysis techniques.Read more: Kolanovic on the Canary in the Coal Mine for Higher Energy PricesOnly Bloomberg.com subscribers can get the Odd Lots newsletter in their inbox — now delivered every weekday — plus unlimited access to the site and app. Subscribe at bloomberg.com/subscriptions/oddlotsSee omnystudio.com/listener for privacy information.

February 03, 2025

Lots More with Matt Levine on MicroStrategy's Infinite Money Machine

Lots More with Matt Levine on MicroStrategy's Infinite Money Machine

Odd Lots

MicroStrategy, a company founded by Michael Saylor, has been buying billions of dollars worth of Bitcoin and the strategy has been successful so far, with its market cap worth more than the value of its Bitcoin portfolio. A new investment strategy is expanding to other companies. Matt Levine of Bloomberg Opinion discusses this in Odd Lots.

January 31, 2025

Get Ready For Another Shock to Housing Affordability

Get Ready For Another Shock to Housing Affordability

Odd Lots

Experts warn of another potential shock to housing affordability due to drop in new apartment development from increased interest rates, labor deportations, skyrocketing insurance costs, and industry consolidation.

January 30, 2025

How Oaktree's Howard Marks Spots a Market Bubble

How Oaktree's Howard Marks Spots a Market Bubble

Odd Lots

Investment expert Howard Marks discusses market bubbles, specifically the 'Bubble.com' note he wrote during the run-up to the 2000 Nasdaq peak and his perspectives on current conditions. Notable for predicting the dot-com bubble and surviving financial crises in 2000 and 2008.

January 27, 2025

AI

Ask this episodeAI Anything

Odd Lots

Hi! You're chatting with Odd Lots AI.

I can answer your questions from this episode and play episode clips relevant to your question.

You can ask a direct question or get started with below questions -

Sign In to save message history