Thursday Nights in AI cover image

Thursday Nights in AI

Episode 10: Dr. Jim Fan, NVIDIA AI

Oct 22, 2023
Dr. Jim Fan, senior research scientist at NVIDIA AI, discusses his career trajectory, the importance of embodied agents in computer vision, challenges in building robust AI products, integration of retrieval and citations into LOMs, the future of gaming with intelligent NPCs, open source tools and reinforcement learning in AI agents, and the significance of research taste and pursuing simple and elegant solutions.
52:23

Episode guests

Podcast summary created with Snipd AI

Quick takeaways

  • Embodying agents in simulations is essential for advancing AI capabilities and bridging the gap towards higher intelligence levels.
  • The future of gaming lies in the development of intelligent NPCs that adapt, learn from experience, and have long-term memory, leading to immersive and unique player experiences.

Deep dives

The importance of embodied agents

Embodied agents, which can control a body and interact with the world, are not receiving enough attention in the field of AI. While most of the compute power is currently dedicated to language models, embodying agents in simulations is a crucial area of research. Simulations offer a safer and faster way to train agents, allowing them to gather a large amount of data and explore complex environments. By focusing on simulations and embodied agents, we can bridge the gap towards achieving higher levels of intelligence.

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
App store bannerPlay store banner