Physical AI Explained: Why Robots Are Finally Learning Like Humans

Robots have been used in manufacturing for decades, but a new wave of physical AI is changing what they're capable of doing. Instead of following rigid, pre-programmed instructions, the latest robots can learn from demonstrations, adapt to changing environments, and make decisions in real time.

In this episode of Today in Tech, Keith Shaw speaks with Andrew Pether, Head of AI Partnerships at Teradyne Robotics, about what's driving this shift and why physical AI could transform manufacturing, logistics, warehouses, hazardous environments, and eventually even the home.

They discuss:

• What "physical AI" actually means

• How robots are learning from people instead of code

• Why AI-powered robots can adapt to real-world unpredictability

• The role of simulation, synthetic data, and real-world training

• How safety works when AI controls physical machines

• Why manufacturing is likely to see the biggest impact first

• Whether humanoid robots are closer than most people think

• Why today's physical AI is both overhyped and genuinely revolutionary

Can AI finally bridge the gap between software intelligence and the physical world? Watch the discussion to find out.