
Sign up to save your podcasts
Or


I sit down with Chris Paxton and Michael Cho, the hosts of the @RoboPapers podcast, to geek out on all things robotics. We dive deep into the massive data gap, the pain of real-world evaluation, and what it will really take to power decentralized production with robots.
Chris and Michael break down the pros and cons of different training methods, why physics is brutal, and why timelines are probably overhyped in the short term and underhyped in the long term.
In this episode, we discuss:
+ The RoboPapers podcast
+ Biggest bottlenecks in robotics
+ Pros/cons of different training methods
+ Michael’s plan to use crypto incentives to gather robotics data
+ The surprising first viable market for robots: Entertainment
+ Why Chris is skeptical of “perfect” robot videos
Enjoy!
Watch on YouTube; listen on Spotify or Apple Podcasts.
Recording date: October 25, 2025
Chapters:
00:00 Meet The Hosts Of Robopapers
04:00 Chris Paxton’s Robotics Journey
06:30 Michael Cho’s Robotics Journey
11:33 Two Main Bottlenecks Holding Robotics Back
16:40 Views On Training Methods: Teleoperation
18:39 Views On Training Methods: Reinforcement Learning
21:02 Views On Training Methods: Learning from Human Video
28:20 Views On Training Methods: Sim-to-Real
30:12 Solving The Training Gap
35:02 Importance Of Robots In The Field
39:47 Entertainment As A Gateway For Robotics
41:27 Why Start the RoboPapers Podcast?
43:43 Most Interesting Areas Of Robotics Research
46:17 Egocentric Data Explained
47:47 Michael’s Crypto Incentive Structure For Data
51:34 Rapid Fire
Follow Jordan on X:
https://x.com/jrwolfe
Links to Chris & Michael’s Work:
RoboPapers Podcast – https://youtube.com/@robopapers?si=NOwQb-rFLeW5mBp1
Chris Paxton on X – https://x.com/chris_j_paxton
Chris Blog - https://cpaxton.github.io/
Chris Substack -
Chris LinkedIn - https://www.linkedin.com/in/chris-paxton-41aba958/
Michael Cho on X – https://x.com/micoolcho
BitRobot - https://bitrobot.ai/
Michael’s LinkedIn - https://www.linkedin.com/in/michael-chung-yeung-cho/
By Jordan WolfeI sit down with Chris Paxton and Michael Cho, the hosts of the @RoboPapers podcast, to geek out on all things robotics. We dive deep into the massive data gap, the pain of real-world evaluation, and what it will really take to power decentralized production with robots.
Chris and Michael break down the pros and cons of different training methods, why physics is brutal, and why timelines are probably overhyped in the short term and underhyped in the long term.
In this episode, we discuss:
+ The RoboPapers podcast
+ Biggest bottlenecks in robotics
+ Pros/cons of different training methods
+ Michael’s plan to use crypto incentives to gather robotics data
+ The surprising first viable market for robots: Entertainment
+ Why Chris is skeptical of “perfect” robot videos
Enjoy!
Watch on YouTube; listen on Spotify or Apple Podcasts.
Recording date: October 25, 2025
Chapters:
00:00 Meet The Hosts Of Robopapers
04:00 Chris Paxton’s Robotics Journey
06:30 Michael Cho’s Robotics Journey
11:33 Two Main Bottlenecks Holding Robotics Back
16:40 Views On Training Methods: Teleoperation
18:39 Views On Training Methods: Reinforcement Learning
21:02 Views On Training Methods: Learning from Human Video
28:20 Views On Training Methods: Sim-to-Real
30:12 Solving The Training Gap
35:02 Importance Of Robots In The Field
39:47 Entertainment As A Gateway For Robotics
41:27 Why Start the RoboPapers Podcast?
43:43 Most Interesting Areas Of Robotics Research
46:17 Egocentric Data Explained
47:47 Michael’s Crypto Incentive Structure For Data
51:34 Rapid Fire
Follow Jordan on X:
https://x.com/jrwolfe
Links to Chris & Michael’s Work:
RoboPapers Podcast – https://youtube.com/@robopapers?si=NOwQb-rFLeW5mBp1
Chris Paxton on X – https://x.com/chris_j_paxton
Chris Blog - https://cpaxton.github.io/
Chris Substack -
Chris LinkedIn - https://www.linkedin.com/in/chris-paxton-41aba958/
Michael Cho on X – https://x.com/micoolcho
BitRobot - https://bitrobot.ai/
Michael’s LinkedIn - https://www.linkedin.com/in/michael-chung-yeung-cho/