I don't really want robot weapons but I would love to see a firing range demonstration with one, just to see what's possible when you get super-human aiming.
I don't see a future for humanoid bots in war. Missiles and drones seems to render anything else redundant. Policing in a few more generations seems possible, at least at a technical level.
It's a shame I can't look at this amazing tech without thinking the same thing. I suspect in reality future soldiers will wish they were fighting these instead of what I assume is more likely in a swarm of flying grenades, though.
idk I'm kind of thinking you could play real-life fortnite with human-controlled robots like these... but then that reminds me of this scifi short movie: https://www.youtube.com/watch?v=1AvyUWUKCw8
A predator drone can fire a hellfire missile from kilometers away and hit a car windshield, killing the passengers with a set of pop-out blades. I don't think humanoid robots are going to be a tipping point.
If I didn't know who Boston Dynamics was, I would have said that this is just CGI. The animations look so fake. I think the issue is that they're aggressively smoothing the motion since mechanical systems don't have the ability to accelerate & jerk as fast as biological systems. Cheap mocap animation has the same problem because the frame rate of the capture system is too low. They interpolate between captured keyframes which results in the wrong velocities in the end.
I felt the same way, but I think it might be because of the context that we're used to seeing advanced robots. Currently, that context is video games and movies, where we know they're fake. In a sense, we're conditioned to see this kind of tech and expect that it is done with CGI, so it makes unbelievable when we see it done for real.
Agree that if anyone except BD put out this video I’d bin it as fake immediately. I still have a hard time saying I believe it - it feels like such a jump from where my mental model of their robots was and CGI is so good at this point that it feels easier for this to be a fake than real.
Same, and I've been working with some of the latest and greatest collaborative and industrial six-axis and SCARA robots. There's definitely a back and forth between innovation and legacy in this industry. For example, FANUC famously refuses to deprecate or drop service for any of its equipment, you'll pay through the nose for some of it but they're willing to keep you supplied with parts, service, and bug fixes for 40-year-old equipment if you've got it. Also, a lot of it is clearly designed for "what's the most we can get an average maintenance tech to understand after a 40-hour certification class", not for programmers and engineers. But with a demo like this, it's obvious that BD is working on a level that the industrial automation competition won't reach for decades.
It might help to compare against video with less editing and effects. Several youtubers are playing with Spots and you can get a better idea of what they're really like.
They're only like what they're "really like" because in those videos they're being operated from behind a layer of abstraction that only allows a few really basic movement patterns (go X direction, at Y speed, within allowed parameters).
I think this new video is trying to illustrate more of the highly dynamic capabilities of the machines that can be used by more specialized operators and software. (And I don't think there are any "effects" exaggerating those capabilities like you're implying, unless you mean things like color correction or the occasional cut to another piece of footage).
It does, but counterpoint: unassisted (no help balancing) bipedal motion has been a holy grail for a looong time, where 4-legged is child's play by comparison. This could mean that just the bipedals are faked.
It would help a lot more to compare footage with real-world people interacting with the bipedal robot.
>It would help a lot more to compare footage with real-world people interacting with the bipedal robot.
Maybe I misunderstood your comment but there are definitely videos of people interacting with Atlas (the bipedal robot) to the point that it's a running joke that they'll remember said interactions. See https://youtu.be/rVlhMGQgDkY?t=83
I would go further than that and say that it's obviously CGI, though I don't know if that actually means it is CGI - it could be, as you mentioned, some sort of post-processing. Or maybe it's greenscreened and composited together?
More than the smoothness, there just seems to be something off about the weight and momentum that you tend to see in high-end, big-budget-movie CGI, but it's hard to put a definite finger on it.
There's a similar effect in old movies when a super-strong character picks up a massive boulder that is actually made of styrofoam or something. You can tell that the mass is wrong based on the way it moves.
I think this is a similar effect but in reverse -- we know the limbs are heavy and aren't made of styrofoam, so the fact that they are moving like they are is uncanny. Our brains aren't used to motorized and hydraulic limbs.
You can get a little bit of this effect watching assembly line arms sometimes, but it's less jarring because they are clearly disembodied, whereas this thing is fairly humanoid.
I was curious so I just timed the jump at the beginning of the song about 15 seconds in.
Assuming the jump is about 30cm, and I stopwatched it at 0.3 seconds from the apex to the ground, you arrive at an acceleration of 9.6m/s^2 which is basically spot on the acceleration due to gravity.
At 0:57 the camera pans to a position where you can see people on the upper floor walking by in the distance. You can see them move as the robot dance, which shows the video is not sped up. You could claim they were composited in (a suicidal move for BD, if they ever tried that), but if you look closely at the windows below, you can see the the camera crew in the reflection, who also move at the correct pace - and that shot would've been hell of a work to fake. In conclusion: this is real-time.
That's a great example. I wonder how much of that is due to the (lack of) the weight of the coffee, vs. not having to compensate for sloshing liquid that could spill if you aren't careful.
Or even in modern movies with CGI-enhanced characters.
Gollum moves as if his enormous head was empty, because the actor doing the motion capture didn't wear a heavy helmet.
In "Rust and Bones", Marion Cotillard plays a character that loses her legs after an accident, and she has her legs removed by CGI, but there are scenes where she's carried by Matthias Schoenaerts that look unnatural, out of balance.
I'm inclined to agree with this after bumping the video up to max quality. I do believe that this is real, and you can see from 0:51-1:05 that there's some reflections going on that would be a lot trickier to fake. But this footage has a similar quality to what I see in a lot of heavily processed drone footage, where it just doesn't look like the same way I see the world.
Also watch their feet very closely at max resolution.
There's a lot of shock rattling and wobbling I've never seen in any CGI robot animation, not to mention they are leaving visible indentations in the foam floor. If it was CG they really went all out.
I think what happened is they used a fancy camera with a big lens and a 30 FPS frame rate, giving a very high quality image reminiscent of Hollywood films, but they set a faster shutter speed than Hollywood generally uses, giving less motion blur than is normal in high quality footage like this. The relative lack of motion blur despite the low frame rate gives it a jerky appearance similar to sped up or timelapse video, or animation. Combined with the slightly odd motion of the robots, it has an uncanny look.
It's kinda funny because most robot demo videos are sped up and/or CGI, but this one isn't.
75% looks most similar to the previous Atlas videos. Also the movement of the people upstairs behind the glass or the reflections of the cameraman confirm that it has been sped up.
I don't know how fast you walk, but at 75% those people are definitely in slow-motion. I could see maybe 90%, but 100% looks completely natural to me as far as their movements are concerned. (Personally, I want to see a jump at 90% to judge, not someone walking.)
A simple test shows this isn't more realistic. Gravity during the jumps are off, and people walking upstairs is odd. It looks like it was filmed in realtime.
I think it's clearly recorded in-camera, and I think real camera footage - the reflections in the (plexi?)glass screens show the robots and people holding cameras. I'm not even seeing any obvious indications it's been denoised.
As someone who works in the VFX industry, faking reflections on wobbling/flexing glass screens (static ones are easy) is actually pretty difficult to do right, and I doubt they would have added the camera operators if they were faking it.
Some of the aspects making people think this is CG are the diffuse lighting (no obvious hard shadows or occlusion), very clean robots, "weird" (i.e. not normal, but I believe it's honest for the robots) motion, and lack of sound.
You can see the vibrations of the plexiglass walls and the camera crew reflected in them vibrating in amazing detail. You can also see people walking around in the background in normal speed. It doesn't look like it's accelerated, or I would say no more than 10%. Also given the visible but subtle marks the robots leave on the mat, this is either next level CGI and compositing or next level robotics, and BD are not a next level CGI company.
I think it's absolutely fascinating that so many people, here and especially in the YouTube comments, are thinking this is CGI, when doing this with CGI would be much harder than doing it for real.
Also, if you've followed Boston Dynamics for a while, what this video is showcasing isn't that much of a leap forward, they're just iterating away on their robots, adding capabilities and movement patterns.
sound would help, and maybe someone can dig up the real audio version if it is available. because hearing those engines roar and servos buzz, you appreciate there is indeed more going on than meets the eye
I think it comes from the use of hydraulics.
Biological muscle is way less linear (it works more like a second-order control system, where things often overshoot their target before dampening).
But one part of it may also be algorithmic: the robot’s Kalman filter (?) may correct for errors from sensor fusion much faster than a human would.
I would love a more informed opinion than mine though.
This video has very little motion blur compared to what you normally see in high quality 4K 24/30 FPS footage (e.g. Hollywood movies). That contributes to the uncanny look, giving it an "animated" or "sped up" feel. Which is funny because most robot demo videos are sped up, but not this one.
To me it looks like claymation. But of course I can't really explain what it means for something to "look like claymation" to me. Is it that the movement is too isolated? Not isolated enough? Both of those things at the same time in different ways?
I agree but I chalked that up to motion of extremely rigid bodies under large torque.
Watch the reflections on the plexiglass while they are moving. Obviously they could have been composited in but there's definitely a live portion of the shot.
Hydraulics make robot motion look fake? Have you watched real robots move before? I've never once watched a robot move and thought to myself "that looks fake".
Oh, man. Hard to watch the sequence starting at 1:56 without wondering when the robot is gonna take that guy's hockey stick and beat him to death with it. Do you want a robot uprising? Because this is how you get a robot uprising.
It can't just be hydraulics though, because to me the most telling moments that look fake are when the robots are jumping - they have too much hang time/too much uniform velocity for the relatively small amount of vertical distance they get. In particular, see the initial jump at 15 secs in the video. It's like the robot is floating instead of jumping.
I had that same distinct impression about the jump at :15 the first time I watched, but watching it again i don’t. I think it has something to do with the leg motion while airborne. If you just focus on the torso it seems normal.
I think it looks fake because the movements are so smooth and fluid.
On a dancing person you see all the slight adjustments to keep balance that make it feel natural, I believe these robots do it too fast for us to notice and so it seems like CGI where keeping balance isn't a thing.
I'm looking at the jump at 2:09, and all I'm seeing is a super wobbly robot trying very hard to keep its balance?
Look at the pelvis movements at 1:15 to 1:25! You can clearly see it jerking around so the robot keeps its balance.
None of those movements are smooth or fluid, it's all jerky jittery and unnatural, which is exactly what you'd expect from actual robots. If it was motion-capture, you wouldn't get those movements, you'd get smooth human movements instead?
I'm so fascinated by how this video triggers people's "it's fake!" senses, when it's clearly real.
For me, the front toe taps are especially fake-looking. It’s hard to pin down exactly without going frame by frame, but to me it feels like it‘s a combination of the lighting and the way the foot “rests” on the floor (which should probably be deforming since it looks like dense foam?)
Almost as though they had to edit out some balance-assisting device.
Also: why are the people at the right speed, but it feels like the robots are sped up? Them being sped up would be fine, but if the two are at different speeds then it’s a clear indicator that it’s composited.
It looks like 'fake' and animation because it _is_ an animation. The dancing was likely motion captured and/or hand animated and superimposed onto the robots basic balancing control system.
I'm not sure what you are saying here. I think the intent is to differentiate between the "optimal" robotic control dynamics which is designed for safety, efficiency, and to respond to unexpected situations and "dancing" which is intended to either mimic human motion or some anthropomorphic version of it.
So it is "animation" of a physical sort designed to make it look like the robots are responding to the music naturally, rather than programmatically.
This gave me the warm fuzzies and I’m not sure how I feel about that.
If I already feel a twinge of empathy for some soulless robots, where are we going to be a few years down the road when this phenomenon is more actively exploited?
I think he meant something of the sort of exploding drones are deadly and don't require that level of AI. and I guess if one of these bots ran towards you, you wouldn't be reluctant to dance with it, while it inconspicuously arms it's self-destruct mechanism. the future is here! STILL NO FLYING CARS.
Why would these things need to be fully autonomous to be scary? The fact that there might be a human operator in the loop would be cold comfort with one of these things hauling ass towards me...
That makes no sense. I would be more worried about it failing like a self driving car would. A malfunctioning robot could set your house on fire through an electrical short even if it is 100% trustworthy and harmless to humans.
Once I am scared of humans wanting to kill me why would I care if a robot with human like intelligence wants to kill me? It's the same thing at that point.
It pushes the ultimate decision higher up the chain of command, for one.
If a general orders a platoon of soldiers to commit a war crime, the soldiers still ultimately have to decide whether to pull the trigger. There's conscience and self-preservation at play. The robots don't have second thoughts.
We tend to think that the first widely deployed military drones are going to be like the movies: big bipedal humanoid robots kicking down doors, armed with machine guns and flamethrowers.
In reality, it'll probably be swarms of tiny quadcopter drones, each with a gun and only a few bullets (and/or some sort of self-detonation ability), shooting out the glass windows and flooding into a house.
There's obviously some AI involved in this scenario, (navigation, mapping, facial recognition), but you definitely don't need to balance a huge machine on two feet to wage an effective drone attack.
You don’t need true AI to make this thing deadly or oppressive.
You don’t need an AI that ponders the meaning of life for you.
You just need a smart Decision and Control System (DCS).
Paired with additional technology for facial and gait recognition, obstacle avoidance, path planning, and you have all the hallmarks of a rudimentary hunter-killer bot.
What Boston Dynamics cracked here, is the physical control mechanism to allow these robots to live in our world.
Next, they need to build the brains, to allow this machine to move independently and autonomously.
Why would anyone be afraid of these? You should be afraid of stealth drones three miles up with air to ground missiles. They are unstoppable without electronic countermeasures.
Its a bit harder to control how much collateral damage occurs with A2G missiles. Put a rifle on a humanoid robot or small explosive on a drone like in Slaughterbots and that’s much more terrifying.
It isn't, really, because you can at least hope to avoid the slaughterbot. You can't escape that which you can't see. A high-altitude UAV is the nightmare, whether it shoots missiles or bullets. We're just used to them (but not being a target of them), so it doesn't feel so scary as imaginary close combat with a robot.
(To this day I remember this one thing about US drone strikes in Pakistan I read in an article many years ago: the people living there became literally afraid of the blue sky - as the drone strikes tended to happen during good weather. That's terror.)
That's a problem. People are not terrified of AGM carrying stealth drones, but they should be because they are already here and in use around the world. Nothing can stop it, no tank, aircraft, rocket, armor. Probably the only realistic defense is an iron dome type system to shoot the missile down. The US deployed a hellfire AGM with retractable swords instead of explosives they are so confident they can land it on a single person or vehicle.
Biped cyborg killer is just not a realistic weapon. The slaughterbot attack on Maduro in 2018 shows that technology is not very mature yet.
Of course they did. The DoD puts enormous resources into publicity stunts. They sponsor movies left and right, do tons of advertising and propaganda, and provide early-stage commercialization funding for consumer products to their contractors. They know BD robots look like murderous war machines, so of course they want to spend some money to make the public like them.
I've seen it plenty of times too, I think it's just because it had shiny gloss panels which is very easy to do in CG so it felt more like something I'd normally see in CG.
As someone who worked in robotics at CMU in the mid 80s with Whittaker, & Raibert, and Crowley this was definitely super exciting to watch. You've come a long way baby. I remember watching you great, great grandfather hopping around in Raibert's lab in 1984. Yeah, and sort of scary.
> pretty terrifying to think we are soon to have people hunting robots
for the robots, maybe
Animal traffic is a lucrative and it seems as illegal as unstoppable. My bet is that if this robots will deploy in a signifiant number, the people soon will realize that can make a life if they trap, club and dismantle expensive robots for selling metal parts and any valuable cargo that they would carry. Will be also illegal but some will try anyways. Anything electric can be insta-fried with a higher voltage or, more slowly, with a cheap molotov cocktail. Not much different than hunting a bear or deer, maybe more profitable.
Meh. If someone wants to "hunt" you with a robot they can already do so with a UAV (think MQ-1 Predator, not an off-the-shelf quadrocopter). It is an already existing true and tested technology. Large governments around the world already maintain literal armies to maintain them and efficiently employ them to kill people. The fact that that weapon is not aimed at you or me is merely a political decision. With legged robots a lot more development is needed before they can do the same. I guess what I want to say: These robots are not deadly yet. The kind of people who can turn them deadly can already kill you if they want. I don't see why you should worry marginally more.
Bombs cause infinitely bigger international uproars for good reasons.
Robot assassins that cannot die change the calculus. Governments don't have to risk losing their human assets in order to take out a target. They can just send out a 4-legged drone to prick them with polonium and call it a day.
Ah, thanks for sharing the song. Haven't heard it in a while. I was listening to Chick Corea's 80s material earlier today though. That era had a lot of cheesy sounding jazz fusion that I can't get enough of!
I'm mostly a rock and roll and metal head, but I listened to a lot of Chick Corea in the 80's, like the Where Have I Known You Before album from his Return to Forever band.
These days, I use a Canadian streaming service called Stingray Music. There, I just discovered a channel called "For the fans of Weather Report". It streams all that sort of stuff nonstop. Mahavishnu, Hancock, ...
I agree those robots are both exciting and scary. After I graduated from Princeton (doing cognitive science related to AI and robotics), I hung around the CMU Robotics labs in 1985-1986 -- mostly as a visitor to Hans Moravec's and Red Whittaker's labs. But I also dropped by other labs, including to learn from Ben Brown, who was an awesome mechanical engineer in Raibert's Leg Lab. Raibert's team had done amazing stuff with the first hopping robot and was then getting going with a first bipedal one.
James Crowley had recently left, so sadly I never got to meet him. I think I would have liked him and vice versa, having probably similar interests in personally-assistive household companion robots -- like Silent Running drones. When I was between two housing rentals that did not quite overlap, I slept one night on the floor of his mostly-emptied Household Robots lab next to a running PERQ computer to keep warm (from the cold air conditioning). The crazy things you do when you are twenty... :-)
It was kind of sad for me (for a couple of reasons) to see Raibert's dynamic robots immobile in the MIT museum when I visited there with my kid a few years ago. But I can be glad those robots got a good home there and the recognition they deserved -- rather than, say, be discarded for scrap like the 1942 Atanasoff-Berry computer at Iowa State (since thankfully reconstructed).
CMU's Robotics Institute was heavily supported by defense dollars back then -- even as at least some (or maybe even all?) robotics researchers there then thought giving robots guns was a dumb idea. There were also researchers -- especially in Hans' lab -- who thought robots would and even should surpass humanity as "Mind Children". But I wondered if even that positive aspiration would really work out as well as some hoped. Since then, I've spent a lot of time thinking about the future of robotics and humanity, and how to get more benefits than harms from robotics. It probably would not take much more than a dumb-ish self-replicating anti-personnel robotic cockroaches to wipe out humanity. (Or even smaller self-replicating things like infectious bioweapons...) And even the smartest AIs like in James P. Hogan's "The Two Faces of Tomorrow" novel might wipe out humanity first in their infancy and then perhaps only later regret it. And if robots and AIs are designed mainly by competitive people to give themselves (or their funders) competitive advantages over other people -- whether for military or business reasons -- bad outcomes seem more likely than if other motives guide their creation.
Even as there is no certainty even given the best intentions, as I saw from a simulation I made a couple years later on a Symbolics in ZetaLisp of self-replicating robotics -- who unintendedly turned cannibalistic during the first simulation run (eating their own children). I had only programmed them to simply assemble themselves to an ideal form from nearby parts and then divide. But after they divided, the most convenient source of spare parts was their offspring. I ultimately had to add a sense of "smell" and scent marking of self/child to avoid that. That unexpected total surprise of the (simulated) robots initial behavior -- obvious now in hindsight -- has stuck in my mind every since.
Although, even as I called the fix "smell", perhaps one might think of it as "love"? :-) Related: "Straight Up -- by
Paula Abdul"
https://genius.com/Paula-abdul-straight-up-lyrics
"I've been a fool before / Wouldn't like to get my love caught in the slammin' door / How about some information, please? / ... Do, do you love me? ..."
In any case, it seems plausible that our direction heading out of any Singularity may have a lot to do with our social direction/maturity going into one. Acting positively and pro-actively may be ...
> It was kind of sad for me (for a couple of reasons) to see Raibert's dynamic robots immobile
This hit a nerve. I was at a CMU event with my daughter a few years back - when she was a member of Girls of Steel. My daughter came and told me excitedly that "my robot was here" and brought me over to where the Terregator was on display. I told the roboticist who was there how exciting it was to see it after thirty years, and that it was wonderful to see that it still worked. He said, "Sorry. Actually it's likely not operating two decade. It was placed here with a fork lift."
Thanks for the reply. The Terregator was historic in its own way -- as perhaps the first robust outdoors-oriented general-purpose wheeled flexible mobile testbed with an AC gasoline-powered generator capable of supporting a significant computer with an onboard interactive terminal. I remember walking behind the Terregator with Red's brother Chuck and others in Schenley park as people tested road-following (well, park-path-following) algorithms. It was cool to see on the terminal display what the robot thought it was seeing in real-time. And it was nice to be outside in good weather. If you made that or worked on it, kudos to you! Terregator was a key innovative step forward as a platform that made possible other follow-on work like for Alvan (all leading up to self-driving cars). Well, when it wasn't trying to drive up trees -- which is what an emergency stop is for. :-)
But yeah to learn it was fork-lifted into place, sigh. I also have my own personal (essentially now non-functional) robotics projects from back then and before squirreled away in a storage area, with bittersweet feelings as I look at them now and then in passing when looking for other things.
Somehow I wonder if -- beside the general issue of becoming middle aged and wistfully looking backwards at past artifacts and past times -- if there is perhaps a certain disappointment that robots today are not that much better in the main that back in the 1980s? Even the video from Boston Dynamics for all its awesomeness doesn't actually show legged robots doing anything useful. We still don't have robots as capable or general purpose as B9 from Lost in Space or the walking drones from Silent Running appeared to be when on the screen (even if in real life they were human-operated fakes). Show me Spot folding laundry (like WG's PR2 could do, if slowly) and I might be more impressed.
Robotics was also so expensive to work on in the 1980s -- plus also the AI winter happened -- which all probably contributed to many people with robotics aspirations (like me, especially to do stuff like this: http://www.islandone.org/MMSG/aasm/ ) instead moving into software-only work, working on programs generally about other things than robots.
Robotics is such a different space today though for a new generation with cheap sensors, Arduinos, Raspberry Pis, reliable WiFi, cheap laptops, cheap platforms, disused old Roombas, and so on... I remember in the 1980s how excited Hans Moravec was of the special (and expensive) "Swedish wheels" he had gotten for a new robot being built (Uranus, an unfortunate name for anything, sadly) which provided sideways mobility without computational complexity -- and now I see those wheels on robot toys.
It's kind of surprising though when I think about that how relatively little progress towards useful activities there has been by hobbyists and even professionals given all that personally affordable potential compared to what one might have expected in the 1980s if we has just had more cheap hardware? Still there is FarmBot, ShopBot, 3D printing, cheaper CNC machines, flying inspection drones, robotic surgical systems, easily configurable commercial vision systems for parts inspection, and probably many other useful things. Even somewhat-self-driving cars that don't kill that many people. But somehow I might have expected more a third of a century later?
To be clear, I am talking about crossing some gap between playing with or learning about mechanisms (like FIRST robotics or Girls of Steel, which seem like genuinely useful educational programs) and actually creating something useful and reliable day-in-day out like a dishwasher, or robot vacuum that deals with pet hair, or window curtain closer, or plant waterer, or robot chef, or reliable Lego sorter or whatever. Or ideally one device that can do all of those things.
Willow Garage was a promising (if pricey, with P...
Thanks; that might be fun. You can also email me at the same userid at kurtz-fernhout.com.
For reference (nothing new for you, but others may find of interest) I just found a writeup on the Terregator where the document was started by others in 1984 but finished by Kevin Dowling years later: https://www.researchgate.net/publication/2299052_The_Terrega...
I was very surprised to learn Kevin was leaving CMU RI for the LED lighting company -- given his commitment to robotics for so long. But, aside from interests perhaps changing over time, perhaps it's an example of people in robotics starting out trying to solve some big complex integrated thing of a general-purpose robot (e.g. Uranus, Terregator, Alvan, etc.) and then ending up focusing on various related details (solid-state electronics and sensors for Kevin; simulations and UI software for me).
And of course Mark Raibert's lab at CMU, then MIT, then BD, has (for decades) focused on improving walking. Walking was an area in which incremental progress could be made in contrast to more complex navigation tasks and manipulation tasks that may have seemed intractable in practice (especially in the 1980s). It's not clear to me though how much special insight BD might have into the larger unstructured navigation issue though.
Yet, as every piece of individual robotics technology gets refined by someone, whether walking, 3D scanning, image recognition, reliable mobile power sources, touch sensors, terrain mapping, and so on, we do get closer to someone putting all those things together again in a way we might have imagined in the 1980s but were always disappointed by each component's limitations. Essentially, BD's progress is necessary in terms of removing another excuse for not having amazing robots (i.e. walking is hard, but BD solved it) -- but by itself that is not sufficient to deliver the robots imagined in sci-fi stories (assuming we really still want them).
I can also guess there may have been a lot of falls edited out of that video? :-) For reference, a section of a larger video showing lots of falling walking robots:
https://youtu.be/xEwtM0pKOV0?t=1154
Having linked to that collection of robot falls, I still feel it is at most a matter of time before all sorts of walking robots (and manipulating robots) can reliably do much better in unstructured environments. It's hard to predict exactly how long though -- whether years or decades. I should have been clearer in my previous post that while computer-controlled mobility is increasingly a solved problem like the BD video shows, using mobility effectively when navigating unstructured environments remains an open issue (similar as for manipulating in unstructured environments remains open). That may sound like a subtle distinction perhaps, but in the 1980s (as you undoubtedly know from first hand experience) just getting a computer to control a big motor reliably was a big deal.
As with your point on baby steps, another example is there are now about 200,000 Kiva robots in Amazon warehouses (somewhat structured environments) where automation makes a big difference i...
The important question is who's to blame when they don't.
Right now there's civil liability for the end customer, integrator, and manufacturer (depending on who failed to follow ANSI/ISO robotic safety guidelines) in industrial automation accidents, and criminal liability if the failure rises to the level of negligence. It honestly works pretty well, every site I've ever worked at has been happy to put safety first and throughput second, to a degree that my ordinarily cynical outlook is pleasantly surprised.
But I don't trust the justice system to correctly follow logical reasoning when these things are used for violence. Who's at fault when a desperate soldier straps something to one of these things and sends it off to commit war crimes - the soldier? The brass who put the soldier in that position with those tools and got that entirely expected result? Boston Dynamics engineers and others who built the tools and shipped them with fine print that says "by clicking OK you agree not to violate the Geneva convention with this"? The robot itself, sentenced to run with worn out bearings and low hydraulic fluid in a long prison sentence?
It brings to mind the Nathaniel Borenstein quote [1]:
> It should be noted that no ethically-trained software engineer would ever consent to write a DestroyBaghdad procedure. Basic professional ethics would instead require him to write a DestroyCity procedure, to which Baghdad could be given as a parameter.
Just look at the history of landmines. 164 countries have voted to ban them, 33 have not. surprisingly (or not), at least one of these countries professes to be horrified when low-tech versions of the same technology are deployed against it and considers that to be appalling moral cowardice.
The "number of countries" is a misleading metric for landmine treaty - as it includes all the many countries who expect no wars at all and so they're not actually giving up anything. The 33 include almost all countries who actually expect to fight any wars and/or have serious militaries. Okay, UK, France and Turkey have banned landmines, but USA, Russia, China, India, Pakistan, Israel and Iran haven't banned them, and those 7 matter far more than the 164 signatories of the Ottawa convention for actual banning of landmines in practical use.
And of course, the roadside IEDs that USA saw in Iraq and Afghanistan would still be legal under the Ottawa convention (it does not ban anti-vehicle mines and remotely detonated devices) so you can't really assert that USA is complaining about something they won't ratify themselves.
If we look at the history of landmines, then it's fairly certain that countries are not going to give up a capability that they expect to use, all the universally accepted treaties only ban things that we consider impractical, which do not give a serious advantage (e.g. chemical weapons, which aren't competitive with conventional weapons if you're fighting against a proper military and not just gassing civilians and/or guerillas).
So combat robots will be banned only if (and while) they turn out to be useless and impractical compared to alternatives, or perhaps selectively banned by the countries who weren't going to use them anyway because they can't afford them. If they are actually good for their role, then they will be legal for all the major militaries.
It's a mashup of clips taken at different times (engineer in the background, no engineer, engineer in the same apot again, etc).
The balancing of the first bipedal robot doesn't seem right to my human eyes, but I think it's conceivable that computers controling a robot could manage it. Also, the dual bipedal robots aren't perfectly in-sync which is either very good cgi, or it's real.
Although I believe it's real (this is Boston Dynamics, after all), CGI characters dancing slightly out-of-sync could be achieved naturally if you motion-captured actors performing the dance.
Are you talking about the cuts that start around 1:50? I think it's pretty obvious those are different takes (there are different robots in different positions instantaneously, I don't think anyone would expect that to be physically possible). The same thing is frequently done with human performers as well.
There's no way Boston Dynamics would publish a hyper-realistic fake (setting aside how much effort that would be). It would be very damaging to their brand.
It’s hard to shake of the feeling that those aren’t rendered. However, it’s also kind of sad that this amazing technology goes from one owner to another without a clear push into the real world.
Bipedal robots aren't a great match for battlefield situations because they lack the ability to do proper decision making and are very expensive.
The dogs might be more usable, but even then modern conflicts are not really based on ground war. I don't think those have any significant value over a an actual drone.
It would be far more effective to just add machine guns to UAVs like they do on the Apache [0]. It's simple, proven and effective. There is absolutely zero need for the bullets to be delivered by a crappy quadrocopter. You can send out quadrocopters for face detection and then just shoot people with the machine gun. This reminds me of that stupid missile that shoots blades. It's completely over engineered.
While I don't disagree with you, there is a very real need to avoid collateral damage in 2020. You cannot expect support at home if your drones are murdering a bunch of kids for each "bad guy" that they eliminate.
You want to be able to kill any leader anywhere and let the organization collapse due to internal power struggles. It's more fashionable.
Seems to be a capital intensive business with no clear path to commercialization.
The robots developed by ABB, KUKA, Mitsubishi and similars are very different to what Boston Dynamics is trying to build.
These are impressive robots but their utilitarian value is too low. An industrial robot by ABB with a clear use case in a manufacturing process costs like $80k. An impressive toy robot with no clear use case like Boston Dynamics’ Spot also costs $80k.
If these robots had a clear use case besides carrying heavy stuff, BD valuation would be 5X or 10X their current valuation.
Of course the real value would be to use their research and know-how on these type of high-mobility/low foot-footprint robotics into actual industrial settings and use cases.
Google didn’t have any internal use case for this tech and commercializing the technology as a vendor is way out of their competence and probably not enough value there to justify extra investment in sales, supply chain, etc.
SoftBank was just doing what it has been doing for the last decade which is investing in shiny things.
Hyundai may be the first owner to actually have a clear use case for this tech.
I just wonder to what extent this is pre-programmed. Surely the engineers gave some sort of high level description, and let the robots figure out the rest. Also, it's probably not possible to acount for all the variability of the environment beforehand, so the "recording" played back does not contain every move to the last detail.
That answers the recording part, but is this the abstraction the programers use? Setting trajectories manually would be hard to do while keeping them balanced, especially with moves involving legs.
A trajectory is a series of time-stamped poses, specified in world coordinates. As to where they come from, yes you need tooling to keep from going crazy. At Anki, they used animation industry tools, so I am told.
The solver handles the balance part in real time, using current sensor input.
The poses are programmed, but the balance can't be. At least.. if they would be that would be a significant achievement as well I think. So the robots have target poses they should approximate, but how to get there, and without falling over, that's seriously impressive.
Exactly right. SOTA is to continuously solve a trajectory optimization problem in realtime. Probably at a rate of a fews 10’s of milliseconds per solution.
Powerful people of the wrong kind will love robots. Robots are loyal and will never say no, will never question orders, they can forget anything at anytime. And they can be implemented in a way so that they cannot feel pain or remorse, they cannot be interrogated, cannot serve as reliable witnesses, etc. Shit will happen and it will be noone's fault.
They do not have a DNA, they do not have a face or an accent, they can technically come from anywhere in the world and be owned by anyone or noone.
Will we be ruled by tyrants backed by robotic armed forces? will people be harassed by throwaway robotic private investigators or hitmen? I can see many ways in which this can end up poorly.
There are many potential constructive uses for robots too, of course... but people in power love control, and I would expect robots to be used to implement control related stuff first, before they get put to work to the service of society. Just like everything else...
*and this would be the place to listen to you hapless nerds speculate about animation and the varied aspects of video fakery involved. These things just mean doom to me.
I was going to ask how long these robots can move without a power cord, but then I found this article from 2015 (https://www.bbc.com/news/technology-30919098) that says it's about an hour. If that was in 2015, I'm guessing they're above 2 hours now. Very interesting. If they could rapid-charge, and could just "take a break" for 15 minutes every 2 hours, they could be used alongside humans doing various manual tasks...
In some of their older videos they have Diesel engines powering them, so they definitely have the capability to make some longer ranges. I was unable to find any articles but if you watch the older videos their engines are extremely loud. Definitely not fully electric.
Jesus y'all are so cynical, commenting on how these are war bots and gonna automate everything away, I just think it's an engineering marvel that these insanely complex robotic maneuvers can be done so fluidly now
Yeah, but can you really blame us? As incredible as this is from a technical perspective, it rings immediate alarm bells about the future of war. These have speed and agility and can reasonably go anywhere a person could go. And YouTube entertainment aside, what really is the use case of athletic robots outside of warfare? It's not much of a leap to attach a gun and connect to a remote human operator. Now you can wage both air wars and ground wars without domestic political consequences.
I was born cynical and I'm proud of it, so apologies to Boston Dynamics and others, but these things will absolutely be used as war bots and automation surrogates once the time is ripe.
Even so, yes, these robots are truly engineering marvels and I wholeheartedly applaud the team for their hard work!
I just think that squinting the eyes slightly past the "search and rescue", the war applications of this kind of technology are obvious: the Atlas could do perimeter guard duty even in a muddy, slightly sloped or otherwise difficult terrain, but it could also form hunt and kill squads who need no rest, whereas the hunted person or persons presumably would.
As for automation, it will certainly happen too: the Atlas is humanoid-shaped and once it matures to be more independent and capable with environmental manipulation, it could easily function as plug-and-play automation for many tasks currently requiring a human. Pick berries or fruits, do gardening, deliver things like mail/pizza, and so on.
Do you know how much berry/fruit pickers get paid? or pizza delivery people? It will be a LONG time before it's more economical for a robot to do it I think
It'll happen very quickly when they're good enough. As you hire more fruit pickers or pizza delivery people the next one costs the same or is more expensive, opposite for robots. Especially since the cost will mostly go into prototypes and research.
As soon as self-driving cars really work as well as Uber drivers for example, they're not going to roll out super-gradually just because the very first cars cost billions to develop and $200k a piece.
Good point. I did not know exactly how much fruit pickers make, I supposed it was not very much. According to Google, the California average is 24 k USD a year, 11.49 USD an hour.
However.
A human must rest after, say, 12 hours, but a robot can just keep going even at night, if the battery charging is swift and smart (e.g. hotswap battery). So a robot could work almost 24 h shifts instead of 12 h shifts. So we get a 2x multiplier here.
A robot can be made to work perhaps twice as fast somehow, maybe it just moves faster or has four arms or can carry more or reach higher faster or whatever. So let's add another 2x.
Now we're at 4x, a single robot doing the job of 4 workers. Now we're at ca. 46 USD an hour for the equivalent of the robot's work.
Would such a robot be hired for, say, 30 USD an hour?
Next, I'll put on an MBA hat (which I don't have) and just pull numbers ouf of thin air; GIGO warning is in effect. Real MBAs please excuse the amateur hour.
Assuming the robot rental shop can keep utilization at 80% and sell it non-stop for a month to orchards, this would be 0.8 * 30 * 24 * 30 = ca. 17000 USD a month for such a robot. Let's say they spend roughly 10% in repairs etc., that leaves about 15 k USD a month. If the robot costs 150 k USD, and it can do fruit picking for ca. 6 months in a year, the robot has paid for itself after ca. two years.
My point was that the fruit orchard pays 30 USD instead of 46 USD for the "same" job (i.e. they get 35% off labor costs).
Also, a robot rental shop in this model actually starts to make profit per robot after two years. This is not such a long time, and the robots might actually become tools like tractors or harvesters are, with similar ownership arrangements, e.g. a farm co-op owns the robots and distributes to members for cheap hourly price.
Although there were many assumptions regarding the numbers, I would not quite agree on robots taking a long time to become economical. Also, the savings at the orchard side would be seen immediately, creating robot demand on that side.
Yea, it's very impressive engineering. However there is no intelligence in these machines. It's a clever combination of hard- and software. Takes human intelligence to build it.
Yeah I think that's always been the appeal of BD, they are bleeding edge in terms of mechanics / robotics. The brains are being developed by entirely different companies, but eventually they'll meet one another.
If we can provide adequate resources to everyone without anyone doing any work, it would make sense to switch to an economy wherein we provide adequate resources to everyone without requiring them to work.
Capitalism is a very effective economic system when having goods depends on enticing people to produce them. It's not a system that sticks around once goods and services become fully decoupled from labor.
I obviously missed the boat on replying to this in time, but here it is, for posterity.
I think the belief that we'll eventually get to a state "wherein we provide adequate resources to everyone without requiring them to work" is wonderful, but unfortunately, unlikely to happen.
I'm sure this has been thought out by people much smarter than me, but I see two problems.
First, who decides what "adequate" is? At some point, a decision has to be made on what resources to allocate to a person. Here, in the US, we "let the free market figure that out", which, ideally, means that people are free to negotiate and enter into transactions as they wish, but in practice, is just another way of saying that the government doesn't intervene in decisions unless laws were very clearly broken, and even then, justice is not quite blind. In any case, aside from laws and regulations on what constitute a minimum level of support for people, we do not define what "adequate" means, probably because it's actually a really tough question. The communists, in theory, wanted to appropriate resources "from each, according to their ability, to each, according to their need." Except the people in power were ruthless and corrupt and that system wasn't all that great at incentivizing people. Anyway, good luck with defining "adequate" and then getting people to agree with you on that.
Second, and more difficult, is that the people who control the means of production now will be somewhat reluctant to give that up in the future, so I'm not sure how you'll manage to produce adequate resources for everyone without requiring them to work if someone else still owns the means of production. If you seize the means of production, you're back to communism.
So, I think the main thing that utopian visions like this miss is the messy reality of human beings, that they're born with inherent inequalities and that those inequalities persist over time. Someone, or, probably a corporation, will own the robots of the future. That corporation will control enormously powerful means of production. There is nothing requiring that corporation to pay you for anything, nor for them to provide you with adequate resources.
It can be an engineering marvel and practically useless in the industry at the same time. I do love the dancing though. And the fact they actually went for boogie-like steps and a choreography rather than a more technical presentation.
So what's the main barrier to using these industrially?
Is it reliability? Did this video a zillion takes?
Is it manufacturing? Maybe it's hard to scale them up?
Is it battery life? Maybe it can power itself for just a few minutes?
Is it object manipulation or sensing/world understanding? Maybe it can't apply the right forces to a soft thing or a flexible thing or know what it can step on vs over or what will move or stay still?
Is it just that anything worth automating is worth specializing, and there are better robots for different specific tasks?
We've seen industrial robots for a while now, and these more general robots have gotten really good, so where are they?
In actual factories, which are highly controlled environments, the things called robots are much simpler and more specialized. So you're "Is it just that anything worth automating is worth specializing, and there are better robots for different specific tasks?" point is more or less it.
But also consider, while these bipedal, quadrupedal and wheeled robots can do all sorts of these in isolation, their ability to accomplish things autonomously in the chaotic, unstructured world outside the factory is little-to-none, accent on none. The Darpa Robitics challenge was essentially considered a failure, all entrant failed. Most could not do the "walk to a door and open it" challenge.
Similarly, Boston Dynamics sells their "big dog" walking robot to the military but it is seldom if ever deployed. It's strong and faster than a horse with basically the use case as a pack horse. But well trained pack horse won't injure if you get in front of it and will walk along with the troops on it's own without a guy with joy stick directing it. And the Big Dog needs constant direction.
Basically, robots don't have even the "animal intelligence" needed for real world activity.
And the Big Dog requires a lot of technical experts surrounding it at all times, an energy production system, and if it falls over or gets stuck it will be harder for one or two people to get it back to orientation.
Until there's light weight batteries or fuel cells available, big, long-endurance robots like Big Dog just aren't practical (especially for military use).
Who needs dancing robots? What I mean this video does not show that useful work has been implemented. Probably flexibility of tasks might be a problem. Or there is no problem and we see rollout in the next few years.
It shows that they can fuse IMU and other sensor data and feed it to a trajectory-optimization system and get actuation torques meeting realtime contraints sufficiently quickly to not fall over. Bonus points for smooth motion.
Check out MIT’s Underactuated Robotics class taught by Tedrake. Spring 2018 is all on Youtube, I believe. A Raibert Hopper simulation is one of the first homework sets. Post a link to tbe video of your solution.
A human in India or China can be had for a couple of hundred dollars a month. These robots probably cost hundreds of thousands of dollars to buy and maintain. And they aren’t as adaptable.
I don’t think this is it. US farming is vastly more productive than in India or China. It has less to do with the cost of labor but the nature of labor, ie these robots don’t have a niche.
Repetitive physical tasks can usually be done with existing machinery and industrial robots. To do anything more complex than that, i.e. non-repetitive tasks, even if they could do it physically, they lack the required brain power. They don't follow verbal instructions except very basic ones.
> So what's the main barrier to using these industrially?
In an industrial setting, a purpose-built robot will outdo these robots, or a human, at nearly all tasks. And whatever these things are doing, likely a human can do it better with a little bit of training- that's low capital cost compared to these robots.
Carrying things around? Conveyor belts. Picking things up? Robot arms. Going up and down stairs? Elevators and/or conveyor belts. If your aim is to make money, a general purpose robot is rarely the best choice.
The only real purpose I can see for these things is being able to travel quickly in rough terrain. Military applications, basically. Just picture a small army of these robots, armed with tasers, chasing you through the woods. While dancing.
> The only real purpose I can see for these things is being able to travel quickly in rough terrain
This reminds me of the (alleged) Tom Watson quote: "I think there is a world market for maybe five computers."
These robots will be everywhere in twenty years. The cost and the software aren't there yet, but it's only a matter of time. Flexibility and time to market will beat purpose-built robots in many cases once the cost comes down. Many purpose-built factories will disappear. Local, distributed generalist factories will spring up, staffed by robots. Drones and autonomous vehicles will deliver "hand-crafted" products directly to customers.
With autonomous vehicles poised to replace the most common job in the majority of U.S. states -- truck driver -- we're on the verge of massive systemic changes. That's at best. At worst, with economic and political tensions already high, we'll get violence and bloodshed on a global scale.
> This reminds me of the (alleged) Tom Watson quote: "I think there is a world market for maybe five computers." These robots will be everywhere in twenty years. The cost and the software aren't there yet, but it's only a matter of time.
Hitting the nail on the head here. I've been following Boston Dynamics for the past 10 years and it's just amazing the progress they've made and it's going to get augmented massively with better battery technology and ML/AI. No doubt in my mind these types of robots will be everywhere in 20 years as you say.
> These robots will be everywhere in twenty years.
Doing what? What can a generalist robot do that a task specific robot can't do better, faster, and cheaper?
I'll be happy to be wrong- they're really neat- but I spent 7 years in the Operations tech division at Amazon, working in warehouses especially, and I don't see any need for these robots vs alternative options.
> What can a generalist robot do that a task specific robot can't do better, faster, and cheaper?
Nothing. And everything.
Nothing, because no matter what task the generalist robot is assigned to, it'll always be theoretically possible to design a custom-built specialist robot that's better at that one task.
Everything, because no matter how much better your sock-drawer-organizing robot is at organizing socks, it'll never be able fold my clothes or make my bed or clean my windows like the generalist robot can. I'm not going to purchase thousands of different robots for thousands of different tasks when I can just buy one that will do all of them.
> Is it object manipulation or sensing/world understanding? Maybe it can't apply the right forces to a soft thing or a flexible thing or know what it can step on vs over or what will move or stay still?
Did you notice the robot's hands? Or, you know, lack thereof?
These robots are impressive, amazing and advanced ... but they are demonstrating mobility, not utility.
If you want your robot to really replace people, it needs amazing hands. Also amazing visual processing, but there's a lot of work being done there by the autonomous vehicle people, and I'm not sure who is currently leading the pack on hands.
Lots of interesting answers to your question, but if I may present another:
There is none.
These robots are currently in use screening COVID patients, assessing parts of the Chernobyl plant and working in warehouses. Those are just the applications that I know of.
I think the "barrier to industrial use" was simply that Boston Dynamics wanted to ensure they had a fairly mature platform before scaling up to significant production.
Price. Only the price is the barrier. :)
But it is getting better quickly as we speak.
Spot is 7000 USD. This is cheap enough to be useful in many situations, but certainly not in your average storage facility.
Only if we can't shift society to accomodate the changing job market and take care of people who have been made obsolete. Ideally by enjoying the fruits of our collective labor and ditching the "everyone must spend more than half their lives working" obsession. Call your senator before it's your head on the guillotine, rich HN readers.
Actually, that's not how it works. Consumer inflation is the go to mechanism for keeping unemployment low. It will keep working even with automation. The government merely has to enact effective policies that keep consumer inflation on track.
There is no meaningful distinction between robots and foreign labor when it comes to low employment rates. Both are "not Americans" if your definition of Americans only includes humans that are citizens of the US. Any inequality that would be caused by robots already exists in the form of inequality caused by globalization.
You can solve inequality quite easily with the right policies. Change tax policy for the wealthy or create demand for labor through economic stimulus.
It's interesting how in manufacturing (and other industries with lots of automation) we are trying to move away from humans as much as possible, while Boston Dynamics is trying to replicate humans as much as possible.
Of course there will always be cases where we want to send a robot into an environment designed for humans (ex. firefighting) so there is value in humanoid robots, but it's interesting to me that we are going in two completely opposite directions with robotics. Imo humanoid robots will have some limited uses, but ultimately not be as useful as robots built for specific tasks.
I believe any company with the word 'dynamics' in the name wants to sell military equipment. So maybe for crowd control armed humanized robots work better (psychologically) than tank shaped.
Look at their product lines--Spot and Handle are the furthest things from humanoid. They've also done some static robot arms.
Atlas is their only active humanoid platform, which they've stated is mainly a research platform just for the sake of advancing robotics and pushing the limits.
Ah, so now when someone suggests "Can't we just drone this guy?" I can at least look forward to a bunch of robots performing the (admittedly great) synchronized dance scene from Michael C. Hall's Gamer before executing me.
477 comments
[ 2.9 ms ] story [ 328 ms ] threadI don't see a future for humanoid bots in war. Missiles and drones seems to render anything else redundant. Policing in a few more generations seems possible, at least at a technical level.
Cirque du Robots
Adam Savage: https://youtu.be/-R8wUybrspo?t=2374
Marques Brownlee: https://youtu.be/s6_azdBnAlU?t=249
I think this new video is trying to illustrate more of the highly dynamic capabilities of the machines that can be used by more specialized operators and software. (And I don't think there are any "effects" exaggerating those capabilities like you're implying, unless you mean things like color correction or the occasional cut to another piece of footage).
It would help a lot more to compare footage with real-world people interacting with the bipedal robot.
Maybe I misunderstood your comment but there are definitely videos of people interacting with Atlas (the bipedal robot) to the point that it's a running joke that they'll remember said interactions. See https://youtu.be/rVlhMGQgDkY?t=83
More than the smoothness, there just seems to be something off about the weight and momentum that you tend to see in high-end, big-budget-movie CGI, but it's hard to put a definite finger on it.
I think this is a similar effect but in reverse -- we know the limbs are heavy and aren't made of styrofoam, so the fact that they are moving like they are is uncanny. Our brains aren't used to motorized and hydraulic limbs.
You can get a little bit of this effect watching assembly line arms sometimes, but it's less jarring because they are clearly disembodied, whereas this thing is fairly humanoid.
Assuming the jump is about 30cm, and I stopwatched it at 0.3 seconds from the apex to the ground, you arrive at an acceleration of 9.6m/s^2 which is basically spot on the acceleration due to gravity.
Gollum moves as if his enormous head was empty, because the actor doing the motion capture didn't wear a heavy helmet.
In "Rust and Bones", Marion Cotillard plays a character that loses her legs after an accident, and she has her legs removed by CGI, but there are scenes where she's carried by Matthias Schoenaerts that look unnatural, out of balance.
There's a lot of shock rattling and wobbling I've never seen in any CGI robot animation, not to mention they are leaving visible indentations in the foam floor. If it was CG they really went all out.
It's kinda funny because most robot demo videos are sped up and/or CGI, but this one isn't.
https://youtu.be/fn3KWM1kuAw?t=58
As someone who works in the VFX industry, faking reflections on wobbling/flexing glass screens (static ones are easy) is actually pretty difficult to do right, and I doubt they would have added the camera operators if they were faking it.
Some of the aspects making people think this is CG are the diffuse lighting (no obvious hard shadows or occlusion), very clean robots, "weird" (i.e. not normal, but I believe it's honest for the robots) motion, and lack of sound.
Also, if you've followed Boston Dynamics for a while, what this video is showcasing isn't that much of a leap forward, they're just iterating away on their robots, adding capabilities and movement patterns.
Let’s talk about why it looks fake.
I think it comes from the use of hydraulics. Biological muscle is way less linear (it works more like a second-order control system, where things often overshoot their target before dampening).
But one part of it may also be algorithmic: the robot’s Kalman filter (?) may correct for errors from sensor fusion much faster than a human would.
I would love a more informed opinion than mine though.
Watch the reflections on the plexiglass while they are moving. Obviously they could have been composited in but there's definitely a live portion of the shot.
This is how the robot moves IN REALITY: https://www.youtube.com/watch?v=uhND7Mvp3f4
In that video you linked, look at second 21/22, when it raises the arms in the air: the way the arms stop instantly is very inhuman.
On a dancing person you see all the slight adjustments to keep balance that make it feel natural, I believe these robots do it too fast for us to notice and so it seems like CGI where keeping balance isn't a thing.
Look at the pelvis movements at 1:15 to 1:25! You can clearly see it jerking around so the robot keeps its balance.
None of those movements are smooth or fluid, it's all jerky jittery and unnatural, which is exactly what you'd expect from actual robots. If it was motion-capture, you wouldn't get those movements, you'd get smooth human movements instead?
I'm so fascinated by how this video triggers people's "it's fake!" senses, when it's clearly real.
Almost as though they had to edit out some balance-assisting device.
Also: why are the people at the right speed, but it feels like the robots are sped up? Them being sped up would be fine, but if the two are at different speeds then it’s a clear indicator that it’s composited.
I don't think "basic" is the right word.
So it is "animation" of a physical sort designed to make it look like the robots are responding to the music naturally, rather than programmatically.
I wondered if they covered the bots in a flowing dress if it would sell the illusion even more.
https://www.youtube.com/watch?v=dKjCWfuvYxQ https://www.youtube.com/watch?v=y3RIHnK0_NE
If I already feel a twinge of empathy for some soulless robots, where are we going to be a few years down the road when this phenomenon is more actively exploited?
Second reaction: When's the OK Go video released?
Third reaction: They paid Boston Dynamics engineers to do this?
Once I am scared of humans wanting to kill me why would I care if a robot with human like intelligence wants to kill me? It's the same thing at that point.
If a general orders a platoon of soldiers to commit a war crime, the soldiers still ultimately have to decide whether to pull the trigger. There's conscience and self-preservation at play. The robots don't have second thoughts.
So far they have only bee used to look rather than touch, but I predict that will change quickly, within the next year or two.
The various military groups around the world have been using semi-autonomy in warfare for decades; you don't need 100% generalizable autonomy.
Plenty of missiles in the past worked just fine with little more than silhouette matching and 1 bit cameras.
In reality, it'll probably be swarms of tiny quadcopter drones, each with a gun and only a few bullets (and/or some sort of self-detonation ability), shooting out the glass windows and flooding into a house.
There's obviously some AI involved in this scenario, (navigation, mapping, facial recognition), but you definitely don't need to balance a huge machine on two feet to wage an effective drone attack.
You don’t need an AI that ponders the meaning of life for you.
You just need a smart Decision and Control System (DCS).
Paired with additional technology for facial and gait recognition, obstacle avoidance, path planning, and you have all the hallmarks of a rudimentary hunter-killer bot.
What Boston Dynamics cracked here, is the physical control mechanism to allow these robots to live in our world.
Next, they need to build the brains, to allow this machine to move independently and autonomously.
https://www.healthline.com/health-news/doctors-reanimated-pi...
(To this day I remember this one thing about US drone strikes in Pakistan I read in an article many years ago: the people living there became literally afraid of the blue sky - as the drone strikes tended to happen during good weather. That's terror.)
Biped cyborg killer is just not a realistic weapon. The slaughterbot attack on Maduro in 2018 shows that technology is not very mature yet.
Not sure if its just so unreal, or maybe there's something unnatural in the movement or if some scenes were sped up.
It’s like a totally different series to that last season with vr-gay/not-gay falcon man and Miley Cyrus signing a bad NiN cover.
But yea. The mean robot dogs, just good short form story telling with so little dialog.
I mean, I guess better a robot than a person.. but still.. pretty terrifying to think we are soon to have people hunting robots.
for the robots, maybe
Animal traffic is a lucrative and it seems as illegal as unstoppable. My bet is that if this robots will deploy in a signifiant number, the people soon will realize that can make a life if they trap, club and dismantle expensive robots for selling metal parts and any valuable cargo that they would carry. Will be also illegal but some will try anyways. Anything electric can be insta-fried with a higher voltage or, more slowly, with a cheap molotov cocktail. Not much different than hunting a bear or deer, maybe more profitable.
Robot assassins that cannot die change the calculus. Governments don't have to risk losing their human assets in order to take out a target. They can just send out a 4-legged drone to prick them with polonium and call it a day.
It has the benefit of being in your face and we having been trained by movies and games that robots that look like that are enemies.
https://www.youtube.com/watch?v=GHhD4PD75zY
These days, I use a Canadian streaming service called Stingray Music. There, I just discovered a channel called "For the fans of Weather Report". It streams all that sort of stuff nonstop. Mahavishnu, Hancock, ...
I'm happy John Mclaughlin is still playing. Here's a song from a couple weeks ago you might like!
https://www.youtube.com/watch?v=BZ1Jx1hf0I8
James Crowley had recently left, so sadly I never got to meet him. I think I would have liked him and vice versa, having probably similar interests in personally-assistive household companion robots -- like Silent Running drones. When I was between two housing rentals that did not quite overlap, I slept one night on the floor of his mostly-emptied Household Robots lab next to a running PERQ computer to keep warm (from the cold air conditioning). The crazy things you do when you are twenty... :-)
It was kind of sad for me (for a couple of reasons) to see Raibert's dynamic robots immobile in the MIT museum when I visited there with my kid a few years ago. But I can be glad those robots got a good home there and the recognition they deserved -- rather than, say, be discarded for scrap like the 1942 Atanasoff-Berry computer at Iowa State (since thankfully reconstructed).
CMU's Robotics Institute was heavily supported by defense dollars back then -- even as at least some (or maybe even all?) robotics researchers there then thought giving robots guns was a dumb idea. There were also researchers -- especially in Hans' lab -- who thought robots would and even should surpass humanity as "Mind Children". But I wondered if even that positive aspiration would really work out as well as some hoped. Since then, I've spent a lot of time thinking about the future of robotics and humanity, and how to get more benefits than harms from robotics. It probably would not take much more than a dumb-ish self-replicating anti-personnel robotic cockroaches to wipe out humanity. (Or even smaller self-replicating things like infectious bioweapons...) And even the smartest AIs like in James P. Hogan's "The Two Faces of Tomorrow" novel might wipe out humanity first in their infancy and then perhaps only later regret it. And if robots and AIs are designed mainly by competitive people to give themselves (or their funders) competitive advantages over other people -- whether for military or business reasons -- bad outcomes seem more likely than if other motives guide their creation.
Even as there is no certainty even given the best intentions, as I saw from a simulation I made a couple years later on a Symbolics in ZetaLisp of self-replicating robotics -- who unintendedly turned cannibalistic during the first simulation run (eating their own children). I had only programmed them to simply assemble themselves to an ideal form from nearby parts and then divide. But after they divided, the most convenient source of spare parts was their offspring. I ultimately had to add a sense of "smell" and scent marking of self/child to avoid that. That unexpected total surprise of the (simulated) robots initial behavior -- obvious now in hindsight -- has stuck in my mind every since.
Although, even as I called the fix "smell", perhaps one might think of it as "love"? :-) Related: "Straight Up -- by Paula Abdul" https://genius.com/Paula-abdul-straight-up-lyrics "I've been a fool before / Wouldn't like to get my love caught in the slammin' door / How about some information, please? / ... Do, do you love me? ..."
In any case, it seems plausible that our direction heading out of any Singularity may have a lot to do with our social direction/maturity going into one. Acting positively and pro-actively may be ...
This hit a nerve. I was at a CMU event with my daughter a few years back - when she was a member of Girls of Steel. My daughter came and told me excitedly that "my robot was here" and brought me over to where the Terregator was on display. I told the roboticist who was there how exciting it was to see it after thirty years, and that it was wonderful to see that it still worked. He said, "Sorry. Actually it's likely not operating two decade. It was placed here with a fork lift."
But yeah to learn it was fork-lifted into place, sigh. I also have my own personal (essentially now non-functional) robotics projects from back then and before squirreled away in a storage area, with bittersweet feelings as I look at them now and then in passing when looking for other things.
Somehow I wonder if -- beside the general issue of becoming middle aged and wistfully looking backwards at past artifacts and past times -- if there is perhaps a certain disappointment that robots today are not that much better in the main that back in the 1980s? Even the video from Boston Dynamics for all its awesomeness doesn't actually show legged robots doing anything useful. We still don't have robots as capable or general purpose as B9 from Lost in Space or the walking drones from Silent Running appeared to be when on the screen (even if in real life they were human-operated fakes). Show me Spot folding laundry (like WG's PR2 could do, if slowly) and I might be more impressed.
Robotics was also so expensive to work on in the 1980s -- plus also the AI winter happened -- which all probably contributed to many people with robotics aspirations (like me, especially to do stuff like this: http://www.islandone.org/MMSG/aasm/ ) instead moving into software-only work, working on programs generally about other things than robots.
Robotics is such a different space today though for a new generation with cheap sensors, Arduinos, Raspberry Pis, reliable WiFi, cheap laptops, cheap platforms, disused old Roombas, and so on... I remember in the 1980s how excited Hans Moravec was of the special (and expensive) "Swedish wheels" he had gotten for a new robot being built (Uranus, an unfortunate name for anything, sadly) which provided sideways mobility without computational complexity -- and now I see those wheels on robot toys.
It's kind of surprising though when I think about that how relatively little progress towards useful activities there has been by hobbyists and even professionals given all that personally affordable potential compared to what one might have expected in the 1980s if we has just had more cheap hardware? Still there is FarmBot, ShopBot, 3D printing, cheaper CNC machines, flying inspection drones, robotic surgical systems, easily configurable commercial vision systems for parts inspection, and probably many other useful things. Even somewhat-self-driving cars that don't kill that many people. But somehow I might have expected more a third of a century later?
To be clear, I am talking about crossing some gap between playing with or learning about mechanisms (like FIRST robotics or Girls of Steel, which seem like genuinely useful educational programs) and actually creating something useful and reliable day-in-day out like a dishwasher, or robot vacuum that deals with pet hair, or window curtain closer, or plant waterer, or robot chef, or reliable Lego sorter or whatever. Or ideally one device that can do all of those things.
Willow Garage was a promising (if pricey, with P...
BD is taking baby steps (sorry) towards the consumer market, and that's one reason that I'm rooting for them.
Economic theory may need time to catch up. I have some theories of my own. Let's talk. ShufflePoint.
For reference (nothing new for you, but others may find of interest) I just found a writeup on the Terregator where the document was started by others in 1984 but finished by Kevin Dowling years later: https://www.researchgate.net/publication/2299052_The_Terrega...
Looking at that sparks some more thoughts...
Kevin himself demonstrates an interesting example of robotics as a career path. He was the first CMU Robotics Institute employee -- but ultimately left to work for a LED lighting company, then on to wearable sensors, and now on to 3D scanning: https://www.cs.cmu.edu/news/dowling-receives-alumni-award https://www.topionetworks.com/people/kevin-dowling-54395e99a...
I was very surprised to learn Kevin was leaving CMU RI for the LED lighting company -- given his commitment to robotics for so long. But, aside from interests perhaps changing over time, perhaps it's an example of people in robotics starting out trying to solve some big complex integrated thing of a general-purpose robot (e.g. Uranus, Terregator, Alvan, etc.) and then ending up focusing on various related details (solid-state electronics and sensors for Kevin; simulations and UI software for me).
And of course Mark Raibert's lab at CMU, then MIT, then BD, has (for decades) focused on improving walking. Walking was an area in which incremental progress could be made in contrast to more complex navigation tasks and manipulation tasks that may have seemed intractable in practice (especially in the 1980s). It's not clear to me though how much special insight BD might have into the larger unstructured navigation issue though.
Yet, as every piece of individual robotics technology gets refined by someone, whether walking, 3D scanning, image recognition, reliable mobile power sources, touch sensors, terrain mapping, and so on, we do get closer to someone putting all those things together again in a way we might have imagined in the 1980s but were always disappointed by each component's limitations. Essentially, BD's progress is necessary in terms of removing another excuse for not having amazing robots (i.e. walking is hard, but BD solved it) -- but by itself that is not sufficient to deliver the robots imagined in sci-fi stories (assuming we really still want them).
I can also guess there may have been a lot of falls edited out of that video? :-) For reference, a section of a larger video showing lots of falling walking robots: https://youtu.be/xEwtM0pKOV0?t=1154
Having linked to that collection of robot falls, I still feel it is at most a matter of time before all sorts of walking robots (and manipulating robots) can reliably do much better in unstructured environments. It's hard to predict exactly how long though -- whether years or decades. I should have been clearer in my previous post that while computer-controlled mobility is increasingly a solved problem like the BD video shows, using mobility effectively when navigating unstructured environments remains an open issue (similar as for manipulating in unstructured environments remains open). That may sound like a subtle distinction perhaps, but in the 1980s (as you undoubtedly know from first hand experience) just getting a computer to control a big motor reliably was a big deal.
As with your point on baby steps, another example is there are now about 200,000 Kiva robots in Amazon warehouses (somewhat structured environments) where automation makes a big difference i...
Then the second robot came into the picture. I went LOL.
Then Spot came into the picture, and my heart skipped a beat.
Then the velociraptor on wheels came into the picture, and I just lost it. I think my brain just melted down.
Next stop: Guns. Lots of guns.
Time to be afraid. Very afraid.
I shudder to think what war variant machines will resort to when they are running low on ammo.
Do robots need to follow the Geneva convention?
Right now there's civil liability for the end customer, integrator, and manufacturer (depending on who failed to follow ANSI/ISO robotic safety guidelines) in industrial automation accidents, and criminal liability if the failure rises to the level of negligence. It honestly works pretty well, every site I've ever worked at has been happy to put safety first and throughput second, to a degree that my ordinarily cynical outlook is pleasantly surprised.
But I don't trust the justice system to correctly follow logical reasoning when these things are used for violence. Who's at fault when a desperate soldier straps something to one of these things and sends it off to commit war crimes - the soldier? The brass who put the soldier in that position with those tools and got that entirely expected result? Boston Dynamics engineers and others who built the tools and shipped them with fine print that says "by clicking OK you agree not to violate the Geneva convention with this"? The robot itself, sentenced to run with worn out bearings and low hydraulic fluid in a long prison sentence?
It brings to mind the Nathaniel Borenstein quote [1]:
> It should be noted that no ethically-trained software engineer would ever consent to write a DestroyBaghdad procedure. Basic professional ethics would instead require him to write a DestroyCity procedure, to which Baghdad could be given as a parameter.
[1] http://www.guppylake.com/~nsb/CSCW-ATOMICMAIL.txt
And of course, the roadside IEDs that USA saw in Iraq and Afghanistan would still be legal under the Ottawa convention (it does not ban anti-vehicle mines and remotely detonated devices) so you can't really assert that USA is complaining about something they won't ratify themselves.
If we look at the history of landmines, then it's fairly certain that countries are not going to give up a capability that they expect to use, all the universally accepted treaties only ban things that we consider impractical, which do not give a serious advantage (e.g. chemical weapons, which aren't competitive with conventional weapons if you're fighting against a proper military and not just gassing civilians and/or guerillas).
So combat robots will be banned only if (and while) they turn out to be useless and impractical compared to alternatives, or perhaps selectively banned by the countries who weren't going to use them anyway because they can't afford them. If they are actually good for their role, then they will be legal for all the major militaries.
The balancing of the first bipedal robot doesn't seem right to my human eyes, but I think it's conceivable that computers controling a robot could manage it. Also, the dual bipedal robots aren't perfectly in-sync which is either very good cgi, or it's real.
There's no way Boston Dynamics would publish a hyper-realistic fake (setting aside how much effort that would be). It would be very damaging to their brand.
The dogs might be more usable, but even then modern conflicts are not really based on ground war. I don't think those have any significant value over a an actual drone.
Now a fleet of tiny quad-motor drones with a few bullets? That could work in high-density urban environment an quickly render guerilla warfare impossible. This project comes to mind: https://www.darpa.mil/program/offensive-swarm-enabled-tactic...
Depressingly accurate. Poor kids are far cheaper.
[0] https://www.youtube.com/watch?t=629&v=vMylpfX7_S0
You want to be able to kill any leader anywhere and let the organization collapse due to internal power struggles. It's more fashionable.
The robots developed by ABB, KUKA, Mitsubishi and similars are very different to what Boston Dynamics is trying to build.
These are impressive robots but their utilitarian value is too low. An industrial robot by ABB with a clear use case in a manufacturing process costs like $80k. An impressive toy robot with no clear use case like Boston Dynamics’ Spot also costs $80k.
If these robots had a clear use case besides carrying heavy stuff, BD valuation would be 5X or 10X their current valuation.
Of course the real value would be to use their research and know-how on these type of high-mobility/low foot-footprint robotics into actual industrial settings and use cases.
Google didn’t have any internal use case for this tech and commercializing the technology as a vendor is way out of their competence and probably not enough value there to justify extra investment in sales, supply chain, etc.
SoftBank was just doing what it has been doing for the last decade which is investing in shiny things.
Hyundai may be the first owner to actually have a clear use case for this tech.
And it must also be said that these are preprogrammed maneuvers on a closed course...
The robot equivalent of performance anxiety.
The solver handles the balance part in real time, using current sensor input.
Have they escaped and gone rogue, or are they just sat on the naughty step?
They do not have a DNA, they do not have a face or an accent, they can technically come from anywhere in the world and be owned by anyone or noone.
Will we be ruled by tyrants backed by robotic armed forces? will people be harassed by throwaway robotic private investigators or hitmen? I can see many ways in which this can end up poorly.
There are many potential constructive uses for robots too, of course... but people in power love control, and I would expect robots to be used to implement control related stuff first, before they get put to work to the service of society. Just like everything else...
And then....
;-)
https://m.youtube.com/watch?v=GM3GM299orc
Even so, yes, these robots are truly engineering marvels and I wholeheartedly applaud the team for their hard work!
I just think that squinting the eyes slightly past the "search and rescue", the war applications of this kind of technology are obvious: the Atlas could do perimeter guard duty even in a muddy, slightly sloped or otherwise difficult terrain, but it could also form hunt and kill squads who need no rest, whereas the hunted person or persons presumably would.
As for automation, it will certainly happen too: the Atlas is humanoid-shaped and once it matures to be more independent and capable with environmental manipulation, it could easily function as plug-and-play automation for many tasks currently requiring a human. Pick berries or fruits, do gardening, deliver things like mail/pizza, and so on.
As soon as self-driving cars really work as well as Uber drivers for example, they're not going to roll out super-gradually just because the very first cars cost billions to develop and $200k a piece.
However.
A human must rest after, say, 12 hours, but a robot can just keep going even at night, if the battery charging is swift and smart (e.g. hotswap battery). So a robot could work almost 24 h shifts instead of 12 h shifts. So we get a 2x multiplier here.
A robot can be made to work perhaps twice as fast somehow, maybe it just moves faster or has four arms or can carry more or reach higher faster or whatever. So let's add another 2x.
Now we're at 4x, a single robot doing the job of 4 workers. Now we're at ca. 46 USD an hour for the equivalent of the robot's work.
Would such a robot be hired for, say, 30 USD an hour?
Next, I'll put on an MBA hat (which I don't have) and just pull numbers ouf of thin air; GIGO warning is in effect. Real MBAs please excuse the amateur hour.
Assuming the robot rental shop can keep utilization at 80% and sell it non-stop for a month to orchards, this would be 0.8 * 30 * 24 * 30 = ca. 17000 USD a month for such a robot. Let's say they spend roughly 10% in repairs etc., that leaves about 15 k USD a month. If the robot costs 150 k USD, and it can do fruit picking for ca. 6 months in a year, the robot has paid for itself after ca. two years.
My point was that the fruit orchard pays 30 USD instead of 46 USD for the "same" job (i.e. they get 35% off labor costs).
Also, a robot rental shop in this model actually starts to make profit per robot after two years. This is not such a long time, and the robots might actually become tools like tractors or harvesters are, with similar ownership arrangements, e.g. a farm co-op owns the robots and distributes to members for cheap hourly price.
Although there were many assumptions regarding the numbers, I would not quite agree on robots taking a long time to become economical. Also, the savings at the orchard side would be seen immediately, creating robot demand on that side.
Edit: typo and formatting
Bottom line. Machines. Not Intelligent.
Capitalism is a very effective economic system when having goods depends on enticing people to produce them. It's not a system that sticks around once goods and services become fully decoupled from labor.
I think the belief that we'll eventually get to a state "wherein we provide adequate resources to everyone without requiring them to work" is wonderful, but unfortunately, unlikely to happen.
I'm sure this has been thought out by people much smarter than me, but I see two problems.
First, who decides what "adequate" is? At some point, a decision has to be made on what resources to allocate to a person. Here, in the US, we "let the free market figure that out", which, ideally, means that people are free to negotiate and enter into transactions as they wish, but in practice, is just another way of saying that the government doesn't intervene in decisions unless laws were very clearly broken, and even then, justice is not quite blind. In any case, aside from laws and regulations on what constitute a minimum level of support for people, we do not define what "adequate" means, probably because it's actually a really tough question. The communists, in theory, wanted to appropriate resources "from each, according to their ability, to each, according to their need." Except the people in power were ruthless and corrupt and that system wasn't all that great at incentivizing people. Anyway, good luck with defining "adequate" and then getting people to agree with you on that.
Second, and more difficult, is that the people who control the means of production now will be somewhat reluctant to give that up in the future, so I'm not sure how you'll manage to produce adequate resources for everyone without requiring them to work if someone else still owns the means of production. If you seize the means of production, you're back to communism.
So, I think the main thing that utopian visions like this miss is the messy reality of human beings, that they're born with inherent inequalities and that those inequalities persist over time. Someone, or, probably a corporation, will own the robots of the future. That corporation will control enormously powerful means of production. There is nothing requiring that corporation to pay you for anything, nor for them to provide you with adequate resources.
Is it reliability? Did this video a zillion takes?
Is it manufacturing? Maybe it's hard to scale them up?
Is it battery life? Maybe it can power itself for just a few minutes?
Is it object manipulation or sensing/world understanding? Maybe it can't apply the right forces to a soft thing or a flexible thing or know what it can step on vs over or what will move or stay still?
Is it just that anything worth automating is worth specializing, and there are better robots for different specific tasks?
We've seen industrial robots for a while now, and these more general robots have gotten really good, so where are they?
In actual factories, which are highly controlled environments, the things called robots are much simpler and more specialized. So you're "Is it just that anything worth automating is worth specializing, and there are better robots for different specific tasks?" point is more or less it.
But also consider, while these bipedal, quadrupedal and wheeled robots can do all sorts of these in isolation, their ability to accomplish things autonomously in the chaotic, unstructured world outside the factory is little-to-none, accent on none. The Darpa Robitics challenge was essentially considered a failure, all entrant failed. Most could not do the "walk to a door and open it" challenge.
Similarly, Boston Dynamics sells their "big dog" walking robot to the military but it is seldom if ever deployed. It's strong and faster than a horse with basically the use case as a pack horse. But well trained pack horse won't injure if you get in front of it and will walk along with the troops on it's own without a guy with joy stick directing it. And the Big Dog needs constant direction.
Basically, robots don't have even the "animal intelligence" needed for real world activity.
https://en.wikipedia.org/wiki/DARPA_Robotics_Challenge
And the Big Dog requires a lot of technical experts surrounding it at all times, an energy production system, and if it falls over or gets stuck it will be harder for one or two people to get it back to orientation.
Until there's light weight batteries or fuel cells available, big, long-endurance robots like Big Dog just aren't practical (especially for military use).
Check out MIT’s Underactuated Robotics class taught by Tedrake. Spring 2018 is all on Youtube, I believe. A Raibert Hopper simulation is one of the first homework sets. Post a link to tbe video of your solution.
Battery technology.
In an industrial setting, a purpose-built robot will outdo these robots, or a human, at nearly all tasks. And whatever these things are doing, likely a human can do it better with a little bit of training- that's low capital cost compared to these robots.
Carrying things around? Conveyor belts. Picking things up? Robot arms. Going up and down stairs? Elevators and/or conveyor belts. If your aim is to make money, a general purpose robot is rarely the best choice.
The only real purpose I can see for these things is being able to travel quickly in rough terrain. Military applications, basically. Just picture a small army of these robots, armed with tasers, chasing you through the woods. While dancing.
This reminds me of the (alleged) Tom Watson quote: "I think there is a world market for maybe five computers."
These robots will be everywhere in twenty years. The cost and the software aren't there yet, but it's only a matter of time. Flexibility and time to market will beat purpose-built robots in many cases once the cost comes down. Many purpose-built factories will disappear. Local, distributed generalist factories will spring up, staffed by robots. Drones and autonomous vehicles will deliver "hand-crafted" products directly to customers.
With autonomous vehicles poised to replace the most common job in the majority of U.S. states -- truck driver -- we're on the verge of massive systemic changes. That's at best. At worst, with economic and political tensions already high, we'll get violence and bloodshed on a global scale.
Hitting the nail on the head here. I've been following Boston Dynamics for the past 10 years and it's just amazing the progress they've made and it's going to get augmented massively with better battery technology and ML/AI. No doubt in my mind these types of robots will be everywhere in 20 years as you say.
Boston Dynamics do not and will not use ML/AI in any of their robots.
Doing what? What can a generalist robot do that a task specific robot can't do better, faster, and cheaper?
I'll be happy to be wrong- they're really neat- but I spent 7 years in the Operations tech division at Amazon, working in warehouses especially, and I don't see any need for these robots vs alternative options.
Nothing. And everything.
Nothing, because no matter what task the generalist robot is assigned to, it'll always be theoretically possible to design a custom-built specialist robot that's better at that one task.
Everything, because no matter how much better your sock-drawer-organizing robot is at organizing socks, it'll never be able fold my clothes or make my bed or clean my windows like the generalist robot can. I'm not going to purchase thousands of different robots for thousands of different tasks when I can just buy one that will do all of them.
Did you notice the robot's hands? Or, you know, lack thereof?
These robots are impressive, amazing and advanced ... but they are demonstrating mobility, not utility.
If you want your robot to really replace people, it needs amazing hands. Also amazing visual processing, but there's a lot of work being done there by the autonomous vehicle people, and I'm not sure who is currently leading the pack on hands.
There is none.
These robots are currently in use screening COVID patients, assessing parts of the Chernobyl plant and working in warehouses. Those are just the applications that I know of.
I think the "barrier to industrial use" was simply that Boston Dynamics wanted to ensure they had a fairly mature platform before scaling up to significant production.
Definitely quite a lot - take a look at the floor underneath some of them, shiny patches worn out by constant beatings I guess.
There is no meaningful distinction between robots and foreign labor when it comes to low employment rates. Both are "not Americans" if your definition of Americans only includes humans that are citizens of the US. Any inequality that would be caused by robots already exists in the form of inequality caused by globalization.
You can solve inequality quite easily with the right policies. Change tax policy for the wealthy or create demand for labor through economic stimulus.
Of course there will always be cases where we want to send a robot into an environment designed for humans (ex. firefighting) so there is value in humanoid robots, but it's interesting to me that we are going in two completely opposite directions with robotics. Imo humanoid robots will have some limited uses, but ultimately not be as useful as robots built for specific tasks.
Atlas is their only active humanoid platform, which they've stated is mainly a research platform just for the sake of advancing robotics and pushing the limits.
I suppose it isn't a bad way to go.