The Future of Robotics Is Teams, Not One Super-Robot

What if the future of robotics is not one general-purpose machine, but a team of specialized robots that learn how to work together?

Amanda Prorok, professor of collective intelligence and robotics at the University of Cambridge, joins Brian Heater for a special episode of Automated recorded at the Davos Tech Summit in Switzerland.

Amanda explains how collective intelligence, robot swarms, and multi-agent systems can solve problems that are too complex for one machine. She traces those ideas from ants using pheromone trails to identify the shortest path to food to AI agents that independently taught themselves the positions used in human soccer.

The conversation explores how drones, wheeled robots, and legged robots can coordinate across different embodiments, why diverse robot teams can be more resilient than identical fleets, and how robots can use cameras distributed throughout an environment instead of relying on onboard perception.

Brian and Amanda also examine the growing safety problem surrounding physical AI. Many robots are being deployed because data shows that they usually work, but researchers still cannot fully explain every failure.

Amanda shares the moment she realized she had trusted a humanoid robot beside her three-week-old baby without knowing exactly what it might do. She then explains her lab’s work on remotely detectable robot policy watermarks, which could allow a phone camera to verify the origin of the AI policy controlling a robot.

KEY MOMENTS
00:00 Physical AI is being deployed without real guarantees
01:30 Welcome to Automated from the Davos Tech Summit
03:25 Growing up across countries and languages
04:44 How language changes the way you think
06:54 Learning to code with a computer scientist father
08:18 Why computer science almost lost her
09:22 The moment swarm robotics changed everything
10:59 What robot swarms learned from insects
12:20 Swarms, multi-robot systems, and collective intelligence
13:50 Why mixed robot teams are harder to build
14:33 How diversity can make robots more resilient
15:56 The cross-embodiment challenge
18:03 A blind robot that sees through its environment
19:15 Centralized, decentralized, and hybrid robot systems
21:21 Why physical AI is making robotics blurrier
22:09 The gap between performance and real guarantees
23:51 Learning from robot failures and edge cases
24:24 The AI soccer experiment
27:13 Robots invent goalkeepers, defenders, and attackers
28:21 Why role specialization appears throughout nature
30:18 How AI discovered a human strategy
31:21 Decentralized control and modular robots
33:17 Why she named it the Prorok Lab
35:05 Building a lab where ideas come from everywhere
36:21 Bringing machine learning into multi-robot systems
37:46 What commercial robot deployment still requires
39:10 Trusting humanoid robots around people
40:15 Watermarking robot behavior with a phone camera
Connect with Amanda Prorok
https://www.linkedin.com/in/aprorok/

Learn more about the Prorok Lab
https://www.proroklab.org/

View Amanda’s University of Cambridge profile
https://www.cst.cam.ac.uk/people/asp45

Explore the robot policy watermarking research
https://www.proroklab.org/publications/iclr2026-watermarking/

We’d love to hear from you. Have thoughts or guest suggestions?
Reach us at podcast@automate.org

You can find the transcript and more episodes of Automated at automated.fm

Unlock full access to Automated and explore everything automation.
Subscribe today and leave a review on YouTube, Apple Podcasts, Spotify, and the Automated Newsletter.
https://www.youtube.com/@automatedpodcast
https://podcasts.apple.com/us/podcast/automated-with-brian-heater/id1837762221
https://open.spotify.com/show/60olq6brlBEIJWggx2fMR6
https://www.automate.org/automation/automated-newsletter

You can also find us on:
LinkedIn https://www.linkedin.com/showcase/automated-podcast-by-a3/
Instagram https://www.instagram.com/automatedpod/