Boston Dynamic's latest Atlas robot

Share unrelated electronics stuff, ideas, rants, etc!
Post Reply
User avatar
uze6666
Site Admin
Posts: 4778
Joined: Tue Aug 12, 2008 9:13 pm
Location: Montreal, Canada
Contact:

Boston Dynamic's latest Atlas robot

Post by uze6666 »

Published a couples days ago but I wanted to share with those who did not see it since it's truly amazing... :shock:
https://www.youtube.com/watch?v=fRj34o4hN4II
User avatar
ry755
Posts: 225
Joined: Mon May 22, 2017 6:01 am

Re: Boston Dynamic's latest Atlas robot

Post by ry755 »

Woah, that's cool! I can't even do a back flip like that! :D
User avatar
Jubatian
Posts: 1560
Joined: Thu Oct 01, 2015 9:44 pm
Location: Hungary
Contact:

Re: Boston Dynamic's latest Atlas robot

Post by Jubatian »

Huh, I had seen this a couple of days before, nice post! It is interesting to see the advancements, how we are apparently getting very near to understand the physics of organic motion systems (Even faceplanting that robot looks very much "real" in the sense that it feels like a normal person could act somewhat alike). Guess it doesn't have too much battery life though!

Creepy? For me, nah, it is amazing. The only creepy part could be what we would be using the technology for in the end, but that's like with anything such as nuclear fission. It could power the world, could get us to Mars and beyond (advanced propulsion systems need massive amounts of energy), while of course sadly it could as well serve for intended or unintended destruction. The most obvious probable benefit of such advanced robotics is giving a new life to people with severe handicaps (think of Stephen Hawking for an obvious example), or could serve well in hazardous environments where organic agility is necessary to traverse (such as even cleaning up Fukushima could possibly be a lot simpler if we had such technology, but you could also imagine advanced space exploration, a reliable Boston Dynamic dog could well replace a rover, and it wouldn't get stuck in a patch of loose soil). Still a long way from there, though, but the advancements are sure spectacular!
User avatar
Kilo
Posts: 330
Joined: Wed Dec 19, 2012 3:43 pm
Location: AC, Germany

Re: Boston Dynamic's latest Atlas robot

Post by Kilo »

Ah.. certainly there's a human in it! :D
Like in C3PO :lol:

No, great video. wow, amazing how far they are!
The last digit of Pi is 7! I calculated this, as I counted to infinity yesterday!
User avatar
ry755
Posts: 225
Joined: Mon May 22, 2017 6:01 am

Re: Boston Dynamic's latest Atlas robot

Post by ry755 »

Kilo wrote: Mon Dec 04, 2017 8:22 am Ah.. certainly there's a human in it! :D
Like in C3PO :lol:

No, great video. wow, amazing how far they are!
That reminds me of Daft Punk's Electroma movie, where the robots want to become human. I'm actually listening to Daft Punk while typing this! They make some pretty cool music 8-)
User avatar
D3thAdd3r
Posts: 3175
Joined: Wed Apr 29, 2009 10:00 am
Location: Minneapolis, United States

Re: Boston Dynamic's latest Atlas robot

Post by D3thAdd3r »

Bring on the machines! I waste way too much time reading about this type of stuff, I really am fascinated by all aspects. I believe humans are smart enough to create something, through applied logic and design, better than what was created by unthinking nature. It might be a while, but the progress is accelerating. Especially for these GPUs, the parallel processing is increasing at a crazy rate, and energy efficiency is also improving(per overall performance). Don't get me wrong I am a humanist, but I refuse to believe things just because they feel better. As such, I can't believe/understand any way that grey matter has some "magical properties" beyond physics, that couldn't be emulated(makes it no less amazing, perhaps more so once "magic" is discarded and the true beauty of it becomes known). I think eventually machines will enjoy several advantages in consciousness, with benefits coming from not being developed in primitive group survival situations. Humans seem much scarier to me than AI, because humans(just my opinion, and as a survival mechanism) are very prone to ideology. Not picking on any ideology in particular, but I firmly believe the bulk of human suffering is directly related to such things. I don't think unbiased logic would go down many of those paths "naturally".

That said, humans could be fairly close to designing themselves. For overall efficiency, self healing, self reproduction, flexible fuel sources, etc. you can't beat the human as a general purpose machine(yet). I would estimate there could be some future designs in biology that could be hard to beat in some aspects. The Human Genome is basically sitting there waiting for philosophy/morality to develop answers before proceeding. That is of course very important to get right, but ultimately it is very important to keep accelerating progress at Science as well.

Eh, no offensiveness to be found in there hopefully. That is just my opinions about such things, which ultimately in the grand scheme of things doesn't mean much :roll:

Edit - wanted to say, if it was proven the human brain employed quantum principles in it's operation...then I think machines are screwed for a long time before achieving that.
User avatar
Jubatian
Posts: 1560
Joined: Thu Oct 01, 2015 9:44 pm
Location: Hungary
Contact:

Re: Boston Dynamic's latest Atlas robot

Post by Jubatian »

D3thAdd3r wrote: Sat Dec 09, 2017 7:57 pmEdit - wanted to say, if it was proven the human brain employed quantum principles in it's operation...then I think machines are screwed for a long time before achieving that.
Nope. Even if it employs that, our machinery is already affected just as well. I mean things since long ceased to be ideal digital '1' and '0'. There are errors and error correction everywhere, even in ordinary home computers. Normal CPUs of course still live in this ideal world, at least from the perspective of the software running on them, but complex systems are already well prepared to handle disturbances. Many of the AI technologies especially wander out of the black-and-white digital domain (think of fuzzy logic). In a modern artificial system there could be millions of independent processing units which could be disturbed, could fail any time, possibly some in every second, yet the overall system still runs, and its exact behavior is affected by those disturbances. Think of butterfly effect. You can no longer predict the exact outcome despite that the components are digital.

This is the new world.

Elevated intelligence will eventually happen. If not, that would be because we drive ourself down the drain, not unlike what happened to classical Greek democracy (if you read about that, you could find a lot of similarities to our present day world!). However if we go on, eventually we will cross that point when we will no longer know what the heck is going on. I would say we are very near, even.

Then... Can you remember THIS. From 1970, almost fifty years before.

Today the Colossus won't be having the key to the nuclear launch system. It would (or already does?) have the key to public media! And that's a power even more scary... I believe we could cross a point when even the powers realize that they can no longer pull the plug (or we already crossed that?), and they are merely witnessing the events unfolding, wrinkled shaking hands on the red button which could only doom it all. If the systems would even still cooperate.

An I am not scared of this. I am scared of humanity. That this can happen because of us in the system. Because a mass of human is just a set of numbers in a statistic, because you can control the crowd.

If such an intelligence was created, it wouldn't necessarily be "evil". Maybe it wouldn't even be possible to associate morality to it (the movie above does a good job in this regard, again, that's fifty years before!). Maybe we wouldn't even be able to affix any judgment on it since when it happens, it will be just too late for us. Maybe we wouldn't even notice. Murky area.

I am doing sci-fi, thinking about such intelligences as well and their consequences. Humanity is clearly in a transient state, something very unstable which is guaranteed to become something different in the oncoming centuries, and what's this compared to the time it takes for species to evolve, life to flourish or fall. Possibly this is one reason why in the recent years I see we are getting the more and more afraid of the idea of contacting any intelligent specie beyond our solar system. If they exist, they would fundamentally differ as it is very unlikely we caught anything in the same glimpse of a transient like us.

Creepy, eh? :)

Anyway, just if this popped up. Opinions. If you feel like it is going too far off, OK, I will remove this post.
User avatar
D3thAdd3r
Posts: 3175
Joined: Wed Apr 29, 2009 10:00 am
Location: Minneapolis, United States

Re: Boston Dynamic's latest Atlas robot

Post by D3thAdd3r »

Jubatian wrote: Sat Dec 09, 2017 9:28 pm If you feel like it is going too far off, OK, I will remove this post.
Eh I think it is fair in a random off topic thread, and actually this has turned into a really fun thread for me unless someone really thinks we need to get the "fun police" in here :)
Jubatian wrote: Sat Dec 09, 2017 9:28 pm Then... Can you remember THIS.
I have never seen this. I need to see this.

I do tend to think the machine will beat biology at some point, for everything, including what we consider "creativity and feelings" which I consider a subset of logic, or some blurring of it. I like to play the "Devil's Advocate" against my own beliefs though, just as much as for them, so that is where the biology angle comes from which I don't really like the concept of.

I think it is hard to say currently that much intentional things are done with quantum effects in computing, save for maybe D-Wave in Canada, potentially having Quantum Computing already a "viable product"...or IBM. That I am aware of, only some very high end cryptography stuff makes *practical* use of quantum entanglement yet. It seems like most all things are still aimed at nullifying all QM effects entirely because they are too complex to design a system around(the ever smaller transistor..not getting smaller to avoid the tunneling). But then again it is hard to imagine technology even in 50 years let alone 200. There could very well be some neural network which operates by design with the "random"(this is a whole other subject there...hidden variables I say) QM effects. Actually that is quite more interesting if that is the required solution for some super AI. That quantum programming looks insanely complex to me though for something large scale. Might have to build AI just to program the next one, as maybe it's already out of reach for the human mind.

On the other side, it might be possible to create brains much larger than the couple kilograms that a human can support. Something like genetic programming, where if molecular biology has density/performance advantages, simply make a bigger better brain with interfaces that allow it to control machines optimized for tasks. What can a 1000 kg brain do, who knows?! The machine approach at least seems to avoid all morality issues, which diving into bizarre biological stuff seems potentially wrong if it is something that is considered a living thing(or dive into really dark places of nihilistic philosophy..maybe that is where everything is going). I feel like Science is something above such loose terms as "good" and "evil" though, or even in all cases is something "good" by nature. But damn there are really people who clearly would do stuff that would be considered completely evil by most of us.

Oh man...extra terrestrials. I can't imagine how they could ever reach us in time, or if they have real life technology like Alcubierre drive to actually get here faster than light without ever moving(in reference), then I really wonder why they would bother with us simpletons, and also I hope they would not!! Something like loggers coming into a forest to harvest resources...they don't care much about the consciousness and suffering of the "lowly animals" there versus their "higher purpose" gains, where I would have to think we would be in a similar position as very simple animals against them, if they can reach us...if they existed. Haha, crazy stuff but it is a big universe out there. My guess is the best we ever get is weird alien bacteria, or less likely some hyper sentient machine coming to wreck us :lol:

Oh wow, this is truly a derailed thread, and I am looking pretty guilty as well :lol:
Post Reply