Amazon is forming a new research lab in San Francisco dedicated to developing AI systems that can interact with both digital and physical environments. The Amazon AGI SF Lab will focus on creating AI agents capable of handling complex workflows using common tools like web browsers and code interpreters.
Key points:
- Amazon's new San Francisco lab will focus on AI agents that can perform real-world actions using everyday tools like web browsers and computers
- The lab is seeded by the recently acquired Adept team and plans to hire dozens of researchers from diverse fields including physics and quantitative finance
- The initiative builds on Amazon's existing AGI team's work, including their Nova foundation models, showing a coordinated approach to AI development
The lab’s leaders, David Luan (cofounder of Adept) and Pieter Abbeel, see these new AI agents as a chance to push the limits. “Our initial focus is on several key research bets that will enable AI agents to perform real-world actions, learn from human feedback, self-course-correct, and infer our goals,” wrote Luan and Abbeel in their announcement.
"We're enabling practical AI that can actually do things for us and make our customers more productive, empowered, and fulfilled," Luan explained in a blog post announcing the launch.
Chris McKay
With a lean team and cutting-edge research agenda, the Amazon AGI SF Lab is the latest strategic bet by Amazon to produce breakthroughs that could reshape both AI capabilities and user interactions. The lab will operate with what Amazon describes as "the agility of a startup with the resources of Amazon," pursuing high-risk, high-reward research projects.