How Boston Dynamics makes their robots run, jump and parkour

Started by René, October 16, 2018, 06:21:15 AM

Previous topic - Next topic

N-drju

This company scares the shit out of me with these contraptions of theirs. Even more unsettling is how they usually treat their own creations, kicking these things around, slamming door at them, knocking them over... God damned Skynet-made-real... ??? ??? Time to get a good baseball bat...
"This year - a factory of semiconductors. Next year - a factory of whole conductors!"

Hannes


René

Quote from: N-drju on October 16, 2018, 06:50:58 AM
This company scares the shit out of me with these contraptions of theirs. Even more unsettling is how they usually treat their own creations, kicking these things around, slamming door at them, knocking them over... God damned Skynet-made-real... ??? ??? Time to get a good baseball bat...

What else did you expect? They are trained to deal with real-world situations. :D

N-drju

I think I just don't belong in the contemporary times, you know? I am a very cautious person in real life and a one that thinks a lot (read: too much).

When other people wow over some technology with their eyes glittering, I usually see potential threats and challenges this very technology may pose in the future. Same was true of smartphones and to some extent my fears materialized - people have become zombified with a seemingly endless array of apps that keep you hooked for many hours. Note, that human interaction and social skills have already deteriorated because of that.

I don't know why and for what reason we need such smart robots. Bomb-disarming and hazardous environment automata already exist and I don't think they need any other intelligence other than that of their operators.
"This year - a factory of semiconductors. Next year - a factory of whole conductors!"

René

Quote from: René on October 17, 2018, 01:28:57 AM
Quote from: N-drju on October 16, 2018, 06:50:58 AM
Even more unsettling is how they usually treat their own creations, kicking these things around, slamming door at them, knocking them over...

I think you have a point with the way they treat their robots, and Boston Dynamics has acknowledged that too. Their design is going to be another problem for them if they want to sell those robots. Especially when they are constantly around us and we interact with them. If they look too human, we'll end up in the uncanny valley, or else we'll personify robots and treat them like pets.

https://www.wired.co.uk/article/boston-dynamics-robotics-roboticist-how-to-watch

N-drju

I don't think any being or construct with human or near-human intelligence would like to be treated as a pet. :( Skip to 0:25 mark:

*https://www.youtube.com/watch?v=3JMvs5f0Mks*

My point is also that, even if nothing genuinely bad happens, sooner or later, intelligent robots will be rendered useless anyway... Once they (accidentally?) learn what is the difference between following orders and doing something out of one's own accord... Do you think they will still be willing to serve us? 3 laws of robotics cannot curb intelligence. Whether it's a flesh or a CPU that generates it.

Maybe this is a bit off-topic thought, but it reminds me how Indian children were "amusing creatures" to European settlers in North America. This was mostly harmless and caused by human ignorance but a scary approach nonetheless.

Or the other way round - imagine a race of aliens that have digitalized themselves and have a superior technology and knowledge compared to ours. How would anyone on this planet feel if they tossed slices of bread and fruit on walkways, laughing that "These critters have flesh eyes and hair! Can you believe it? #lol #terranspringbreak #mylittleearthling".
"This year - a factory of semiconductors. Next year - a factory of whole conductors!"

masonspappy

Quote from: N-drju on October 18, 2018, 07:40:34 AM
imagine a race of aliens that have digitalized themselves and have a superior technology and knowledge compared to ours.

Think "Robinette Broadhead"...

WAS

Think about this is we can litterally flash chips and subsystems to monitor that behavior that even the AI itself wouldn't have access too. Just like read-only chips on a PC that govern the boot cycle and behind the scenes work.  Think about a MIPS station. It has a back-controller which monitors the life of the system and can display error information you otherwise wouldn't have access too.

PabloMack