In this episode, Sabrina Halper sits down with Eric Jang. Eric is the VP of AI at 1X Technologies, a humanoid robot company built to benefit society, with investment led by OpenAI. He also published AI is Good for You, a book about the last 10 Years and the next 10 Years of Artificial Intelligence (link to buy attached below). Previously, Eric was a senior research scientist at Google on their robotics team.

Timestamps:

(0:00) - Highlights

(2:53) - Intro

(3:32) - The outdated political system & AI's applications for campaigns7:55 - Debating ideas with AI "champions"

(9:20) - How Meta's VR Quest headset could be a humanoid robot, data engines for AI

(13:20) - Building genuine AI companionship

(18:26) - AI chatbots as sparring partners and improving peoples' social capabilities

(20:45) - Can we achieve AGI through online, text-only data? Or do we need a data engine such as VR, robots, or cars?

(24:20) - How to acquire and label "truth" for LLMs?30:00 - How far are we from AGI? How do you define AGI?

(32:20) - Should we be modeling AI systems after biology and replicating nature-inspired architectures?36:00 - Confronting workforce disruption of humanoid robots, 1X's approach39:00 - Lump of labor fallacy, 10X-ing workforces

(40:10) - Security risks of humanoid robots, high leverage technology = risk

(42:35) - Engineering hardware to mimic the human body, lack of robotics parts

(46:04) - Why is now the right time to invest and build in general purpose robots?

(48:25) - Visualizing a post-AGI future of abundance & utopia situation

(52:27) - What will our relationship to robots look like and which hierarchies will exist? How to engineer an equality dynamic

(59:54) - Criticisms of the effective altruist and AI safety communities

(1:04:20) - Good policies and models for regulation, issues with regulation today