A general “tell” of control strategies is beauty. The more elegant it looks, the more symmetry there is to the behavior. Which means many variables are being reduced to a few via symmetry-exploiting maneuvers.
Conversation
Contrast with the awkward flailing of trying to learn *any* workable functional behavior. Babies and injured pets often learn weird looking maneuvers this way. But more elegant looking ones are either strong attractors everybody will stumble into, or explicitly trained.
1
20
Hardware — solved
Basic control — solved
Maneuver control — solved
Localized deep learning: 2022
Deep learning entire behavior envelope from humans: 2024
Doing it without human training examples: 2026
Learning the unstable/high-risk bits in simulation: 2028
Akira entity: 2030
7
2
39
Replying to
My predictions regarding robotics advance is roughly similar timeline. Do you think independent self-funded research with cheap bots (spot knockoff, robotic arms) could be a valid alternative to big labs and academic path?
1
I am planning to replicate current research and build upon questions here but aligning towards some real world use. youtu.be/l9pwlXXsi7Q
1
1
Replying to
that's what we're doing with the yak rover project... a bunch of us just building rovers with low-cost self-funded research. I've spend about $1000 in parts and tooling for mine so far...
1
1
Replying to
I'm at fork in the road so to speak. Leverage money from a well paying job towards funding this type of experiments. Or go for a research masters in germany, then figure things out by getting into their lab. I would really appreciate your thoughts here if any.
2
I am seeing open communities like eleuther.ai that do scales dl research. Which makes me wonder if independent collabration via internet towards advances in robotics makes real sense. Rather than a impoverished PhD student route.
2
1
Replying to
My grad school experience is almost 20y ago, and conditions have changed, but there's still a lot of value to being around others in a campus environment with lots of serendipity. Online open collaboration is good in a complementary way. They're not substitutes really.
1
1
Replying to
Thanks! Value of serendipity in such spaces was precisely the question that was eating me up. I always knew it was big in back of my head.
The either/or comes from full job +online/free-time robotics might be better because the money might help more towards the goal.
2
Replying to
depends on your age frankly... the younger you are, the more you can afford to forgo money for a couple of years of low-paid grad-student explorations
Replying to
I'm 27, and can fund this masters stint without debt. And I will make sure to get a job in a lab there. Yeah, I guess I can afford this exploration.
1

