AI Summer

Nathan Labenz on the future of AI scaling

42 snips
Jan 27, 2025
Nathan Labenz, host of the Cognitive Revolution podcast and an AI scout, joins to discuss the recent slowdown in AI scaling. He notes that while technology adoption has lagged, significant advancements still occur in model capabilities. Labenz anticipates continued rapid progress, maintaining that we're still on the steep part of the scaling curve. The conversation also highlights AI's potential to discover new scientific concepts, emphasizing the need for a deeper understanding of scaling laws and the complexities within AI organizations.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

Slower Adoption Than Expected

  • AI advancements haven't met initial expectations regarding widespread deployment.
  • Model capabilities, like expanded context windows and tool use, have significantly improved.
INSIGHT

Capability Progress vs. Deployment

  • AI model capabilities have progressed significantly in various areas, including cost, speed, and multimodality.
  • Despite this progress, real-world deployment into applications has been slower than anticipated.
INSIGHT

Scaling Laws Still Hold

  • Despite rumors of development challenges, AI leaders suggest the scaling curve remains steep.
  • Continued progress in AI capabilities is expected.
Get the Snipd Podcast app to discover more snips from this episode
Get the app