@ppitti Happened to me last night too. Checkout this video, it shows how building shortcut roads connecting larger roads actually increases travel time (the chokepoints you talk about). https://t.co/3h2jquhetb
Brightest minds in robotics are working on tasks like cleaning sink, folding laundry, making beds. Are these hard problems?Yes. But do they reflect lack of ambition and disconnect from real-world probs?Yes. Billions poured into chasing LLM moment for robotics that may never come
@animesh_garg What kind of planner here determines grasp approach and grasp timings? Is the grasp triggered only when the error to the desired pose is minimal or is it somehow encoded in the trajopt? And does cuRobo provide the grasp integration too?
@ChongZitaZhang they prolly use adaptive control to estimate object dynamics. Just need to estimate the gravity/weight, other dynamical components are fairly minimal for objects
Go talk to anyone in the military and ask what they’re most afraid of.
Doesn’t matter what they do—infantry, aviation, navy, intel, special ops, logistics, or anything else.
They’ll all tell you it’s this.
I feel a huge sense of whiplash in robotics. On one hand, SOTA "generalist" policies don't generalize to even small shifts in initial conditions, distractors, etc. But there are 100s of millions pouring into startups w/ the same technology. Genuine question: what am I missing?
@philipcball Somewhat related I remember reading H.A. Bethe and V. Weisskopf arguing(in a Science article) on the contrary that solid state theory/biology/phenomenology hardly advances our understanding of the nature.
@philipcball Philip Anderson, a Nobel laureate, coined the expression 'More is Different' to emphasize that even though particle physics is more fundamental, if we scale the system-from micro to macro-new laws emerge that are independent of actual structures that govern them at micro level.