"Move 37" is the word-of-day – it's when an AI, trained via the trial-and-error process of reinforcement learning, discovers actions that are new, surprising, and secretly brilliant even to expert humans. It is a magical, just slightly unnerving, emergent phenomenon only
MACHINE LEARNING
-
Atlas Robot Exercises by Boston Dynamics Engineering
By
–
Atlas #Robot Exercises
— Ronald van Loon (@Ronald_vanLoon) 28 janvier 2025
by @IntEngineering#Robotics #MachineLearning #ArtificialIntelligence #ML #Innovation
cc: @miketamir @chr1sa @karpathy pic.twitter.com/BV60dQW17BAtlas #Robot Exercises
by @IntEngineering #Robotics #MachineLearning #ArtificialIntelligence #ML #Innovation cc: @miketamir @chr1sa @karpathy -
Core Technology Remains Fundamentally Matrix Multiplication Based
By
–
Not really — the core is still just big matmuls
-
RL vs SL: Demystifying Reinforcement Learning’s Core Concept
By
–
Yeah exactly. I get triggered when RL is dressed up in its full rigorous math formalism because it's gate-keeping an essentially trivial core idea. SL:
a token sequence comes from some 3rd party source (e.g. human demonstration), and you just train on it. RL:
you first sample a -
Plasma Simulation and Generative AI: Career Paths in Fusion Research
By
–
Ironically I shifted to fluid dynamics in 2022 (plasma simulation for fusion power) so we moved in opposite directions. Thanks for GigaGAN!
-
Sparse vs Dense Models: Pre-training and Inference Trade-offs
By
–
Its advantages, George, are best understood for pre-training. How it performs in inference remains a less well-understood area. And there may be trade-offs there where "dense" models do better.
-
Enterprise AI Success: Moving Beyond Hype With New Strategy
By
–
Beyond The Hype: The New Playbook For Enterprise #AI Success
by Umesh Sachdev @Forbes Learn more: https://
buff.ly/3YerzIV #ArtificialIntelligence #MI #MachineLearning #Tech cc: @amuellerml @sallyeaves @marcusborba -
AmbiRobotics Launches AmbiStack: AI-Powered Adaptive Palletizer
By
–
.
@AmbiRobotics launches AmbiStack: an AI-based palletizer trained with Reinforcement Learning to adapt on-the-fly to varying box sizes to optimize packing density: https://
businesswire.com/news/home/2025
0128540125/en/Ambi-Robotics-Introduces-AmbiStack-an-AI-Powered-Robotic-Stacking-Solution-for-Warehouse-Operations
… -
DeepSeek R1: How Logic-Based AI Learning Works
By
–
here's everything you need to know about DeepSeek—how it works, and why it matters: how it works instead of relying on human labelers to teach AI what "sounds right," R1 learns through pure logic. when it solves a math problem or writes code, it's rewarded only when the answer
-
Vincent Boucher Achieves Top OpenAI Gym Ranking With AI Agent
By
–
[ The Power of Anticipation 👾 ]
— MONTREAL.AI (@Montreal_AI) 28 janvier 2025
Vincent Boucher: #1 worldwide on the OpenAI Gym.
OpenAI Gym: https://t.co/0XKJf64cZf
YouTube: https://t.co/W4teQShmL6#ASI #AIAgent #AGIALPHA #MONTREALAI pic.twitter.com/aCkRkxi6Yh[ The Power of Anticipation ] Vincent Boucher: #1 worldwide on the OpenAI Gym. OpenAI Gym: https://
web.archive.org/web/2016111316
5326/https://gym.openai.com/envs/Pong-ram-v0
… YouTube: https://
youtu.be/yoETTznoMUA?si
=JRZShLgF3t62MYqe
… #ASI #AIAgent #AGIALPHA #MONTREALAI