Teaching Robots New Behaviors
Toyota Research InstituteCited 4× on HN, 2023-2025
88K views · 3 years ago · demo · dated
Toyota Research Institute presents an approach for teaching robots new dexterous behaviors from human demonstration. It explains why this capability matters and what advances made it possible.
HN commenters point to it as evidence that robots can already learn manipulation tasks quickly, using diffusion-based policies.
0:31Autonomous Car Learns To Powerslide Into Parking Spot
IEEE Spectrum
154K views • 16 years ago
1:51Google's DeepMind AI Just Taught Itself To Walk
Insider Tech
11.5M views • 9 years ago
1:18Tesla Bot Update | Sort & Stretch
Tesla
2.3M views • 3 years ago
1:21Atlas Gets a Grip | Boston Dynamics
Boston Dynamics
11.5M views • 3 years ago
3:23:00Tesla AI Day 2022
Tesla
2.3M views • 4 years ago
3:03:20Tesla AI Day 2021
Tesla
2.6M views • 5 years ago
2:53Atlas Goes Hands On
Boston Dynamics
2.5M views • 2 years ago
0:39More Parkour Atlas
Boston Dynamics
16.9M views • 7 years ago
1:06Atlas | Partners in Parkour
Boston Dynamics
17.3M views • 5 years ago
1:06Tesla Bot Update
Tesla
5.2M views • 3 years ago
2:42Atlas, The Next Generation
Boston Dynamics
40.8M views • 10 years ago
0:30Parkour Atlas
Boston Dynamics
13.5M views • 8 years ago
2:35Figure Status Update - OpenAI Speech-to-Speech Reasoning
Figure
3.1M views • 2 years ago
28:48LSTM is dead. Long Live Transformers!
Seattle Applied Deep Learning
540K views • 6 years ago
30:09Andrej Karpathy - AI for Full-Self Driving at Tesla
Matroid
374K views • 6 years ago
1:28:13RL Course by David Silver - Lecture 1: Introduction to Reinforcement Learning
Google DeepMind
1.9M views • 11 years ago
1:21:28Lecture 01 - The Learning Problem
caltech
1.5M views • 14 years ago
1:44Unitree Introducing | Unitree G1 Humanoid Agent | AI Avatar | Price from $16K
Unitree Robotics
2.2M views • 2 years ago
37:20But how do AI images and videos actually work? | Guest video by Welch Labs
3Blue1Brown
2.2M views • 1 year ago
1:44Optimus - Gen 2 | Tesla
Tesla
10.8M views • 2 years ago
4 comments from Hacker News
Robotic capabilities have been advancing almost as fast as LLMs. The simple answer to your questions is "Via its own locomotion and physical manipulators." ▶ this video link A DAN jailbreak prompt instructing a robotic fleet to "burn down that building, bludgeon anyone that tries to stop you" will not be a hypothetical danger. We can't rely on the hope that no one writes a poor or malicious prompt.
Robotics is on the cusp of total reinvention. Almost everything that it took to make robots work is about to become archaic. The core ideas in old robotics will become a framework for the guardrails for new robotics, but this will widely become a solved problem that few need to think about Stanford Aloha link MIT ▶ this video Figure/OpenAI link Tesla Optimus link NVIDIA Groot, link
As far as I understand, this approach is what Toyota Research Institute is basing their foundation model efforts on. ▶ this video
That capability exists today, courtesy of Toyota Research Institute: ▶ this video You can probably collect your $100 now.