Robotics Made Simple: Playing with LeRobot and SO-101
Exploring Hugging Face’s LeRobot with the SO-101 — from unboxing to first experiments.
All of my long-form thoughts on programming, leadership, and more, collected in chronological order.
Exploring Hugging Face’s LeRobot with the SO-101 — from unboxing to first experiments.
Step-by-step guide for getting Isaac GR00T working on an RTX 5090: PyTorch ≥ 2.7, CUDA 12.9, and Flash Attention 2.
The RTX 5090 isn’t detected out‑of‑the‑box on Ubuntu 24.04. You need to install drivers from the NVIDIA website directly.
Springtime in Boston + cutting‑edge robots = one epic show. Check out my laid‑back photo tour and key takeaways.
While some may argue that overall unemployment is rising - and that we already have enough workers - this assumption does not hold true across all industries. In fact, certain sectors are experiencing labor shortages where automation and robotics are the only path to scaling operations.
The expo floor was buzzing with passionate founders and business professionals, and AI startups were everywhere! I believe this AI surge will push companies to discover new areas and methods for its application, leading to significant growth in robotics and IoT in the coming years.
Whenever I start a conversation about electric vehicles (EVs), the first thing people tend to talk about is the environment – fighting global warming, cutting down on pollution, and all that. Dont't get me wrong, those are great points. But honestly, I think progress in the EV industry is needed to go beyond being eco-friendly, even though EVs aren’t as "eco" as some may think (subject for a separate conversation).
My first recommendation to non-technical founders is simple: find someone with strong IT expertise to join your team as a Chief Technical Officer (CTO). The right person in this position can guide you through all those tricky technical decisions, and together you will turn the idea into reality.
The term "AI" itself seems too convincing for many, creating an illusion that it can handle almost any task and even better than humans.
While today most people focus on which company will develop the most advanced LLM, I want to look beyond this horizon. As someone deeply involved in the IoT and helping companies connect devices to the cloud, I've been pondering the future role of IoT in the world of AGI.
LLMs, they're like a mix of super-advanced data compression and search algorithms. They try to predict the next "token" based on the training data and prompt you give them. So, when you ask them to count something, they don't really count - they just make an educated guess. They simply can't count.