Is AGI Coming? The Role of Data, Evaluation & Human Judgment

Thu, Nov 19, 05:00 PM - 05:30 PM UTC

Speakers

Vineeth VelmuruganVineeth VelmuruganRobotic Learning LeadEncord
Rajat KumarRajat KumarApplied Science Manager & Science LeadAmazon
Siddharth KothiyalSiddharth KothiyalSenior Machine Learning EngineerNvidia
Brad BycraftBrad Bycraft​Product LeaderApplied Intuition

Register now

Fill out your details below and we'll send through the dial-in link. If you can't make it fill out the form and we'll send you the webinar recording.

Live panel on general intelligence in the physical world.

If general intelligence is coming, it will have to work in the physical world: perceiving messy environments, handling edge cases no one labeled, and acting safely around people. That makes robotics and autonomous systems the hardest test, and it makes data and evaluation the real bottleneck.

In this 45-minute panel, engineers building perception, robotics, and foundation models at Amazon and NVIDIA get into:

  • Is AGI actually coming, and is physical AI the real test?
  • Why scaling data for embodied systems is harder than scaling text.
  • How you measure progress when models leave the benchmark and enter the world, and what good eval looks like for robots and AVs.
  • Where human feedback, annotation, and review still matter, and how to scale them without becoming the bottleneck.

SPEAKERS: 

  • Rajat Kumar, Applied Science Manager & Science Lead, Amazon: Rajat is an Applied Science Manager and Science Lead at Amazon, focused on robotics, after earlier work on Alexa’s language understanding models and LLM research at Amazon AGI. He holds a PhD in Biomedical Engineering from Stony Brook University.
  • Siddharth Kothiyal, Senior Machine Learning Engineer, Nvidia: Siddharth is a Perception Engineer at NVIDIA, building real-time deep learning and computer vision systems for autonomous vehicle perception. He holds a Master’s in Robotics from Johns Hopkins, where he researched vision-based path planning for drones, and previously worked at Samsung R&D. 
  • Brad Bycraft, ​Product Leader, Applied Intuition: Brad is a Product Leader at Applied Intuition and former VP of Engineering at Sanctuary AI, where he developed several generations of humanoid robots and deployed the first commercial humanoid robot. He studied Engineering Physics at the University of British Columbia.

Get the data right.

300+ of the best AI teams in the world use Encord.