Artificial Intelligence: More Than Just Computing Power Counts

Eulerpool Research Systems Sep 24, 2025

Takeaways NEW

  • Organizational structure and resource utilization are crucial for innovation.
  • Computing capacities are important, but not the only success factor in AI.
While the focus of many technology companies is on providing powerful graphics processors, Jason Kwon of OpenAI highlights a differentiated perspective. In the "Auren Hoffman" podcast, the Chief Strategy Officer emphasizes that computing capacities are fundamental for the AI industry, but do not represent the sole success factor at the organizational level. According to Kwon, it's more about how resources are used and deployed to tread innovative paths. Three factors are crucial: scarcity, the selection of bets, and organizational structure. Scarcity can foster innovation by forcing more deliberate decisions regarding resource use. Choosing research directions and the right timing to pursue projects intensively or make a course correction gives a lab an advantage. Furthermore, Kwon emphasizes the importance of the right structure to act successfully in the long term, and refers to the necessary instinct for the right projects. On a national level, Kwon still sees computing capacities as central, particularly because of the diverse experiments they enable. OpenAI is known for its immense demand for computing power. CEO Sam Altman recently announced that the company is testing new functions by utilizing significant computing power. Chief Product Officer Kevin Weil stressed that every new GPU that comes in is immediately put to use. In comparison, Elon Musk's xAI also relies on powerful computing infrastructure, while Mark Zuckerberg from Meta announced an increase in spending on GPUs to create a competitive advantage for researchers.

Eulerpool Markets

Finance Markets
New ReleaseEnterprise Grade

Institutional
Financial Data

Access comprehensive financial data with unmatched coverage and precision. Trusted by the world's leading financial institutions.

  • 10M+ securities worldwide
  • 100K+ daily updates
  • 50-year historical data
  • Comprehensive ESG metrics
Eulerpool Data Analytics Platform
Save up to 68%
vs. legacy vendors