Boston Dynamics announced their new and improved Atlas robot - the next step in commercializing their humanoid robot development. We look at how it compares to Tesla's Optimus robot. #robotics #robots #tech #bostondynamics #tesla
Tesla's AI driven software stack (derived from their autopilot work) is extremely impressive but their hardware is only about 2 years old. Boston's hardware is extremely mature but their software is still fairly basic and doesn't incorporate nearly as much AI. The real question is which one is more important -- great hardware or great software?
more like a bugatti and toyota. Boston Dynamics does have better performance but it is certainly gonna be hard to mass product the Atlas. Optimus is built with the purpose to mass produce and actually impact the industry.
@@WillyKo-s6e Mass producing optimus might be the long term goal, but the current version isnt even close to being mass production ready. The entire robot is made out of cnc parts, which are both extremely expensive and extremely time consuming to make.
that's just one thing, everything in Boston turned out to be scripted and scripted and yes, they went bankrupt and the Koreans bought them out...Hyundai Carl!
There’s just no comparison between Boston and Tesla. Just look at Atlas, it moves like a living thing, and that’s the old retired one. Boston is like 10 years ahead of anything else. For for them to retire a robot that’s that sophisticated they realy must’ve made something mind boggling.
Yeah lmao Tesla can barely even walk, Boston dynamics is just on a whole different level with how smooth their robots move, like just compare boston's robot atlas literally just jumping to the Tesla robot trying it's best to walk, not even a fair comparison.
@@thomaslilly5834 That's not the point. The point is that people are touting Optimus as competition to Boston Dynamics's Atlas, which is like saying a Homemade Go-Cart is Competition for a Ferrari.
@@Breakdown5297 It is actually kind of ironic that by saying "that's not the point", you are missing the point completely. Again: Boston Dynamics had 30 years to develop their line of robots. OF COURSE they are much better, and years ahead. THE POINT IS that Tesla (and other companies) only needed 3 years to develop a product that, while not being there yet, is catching up with incredible speed. So maybe in a year or two, they may start being competitive.
@@thomaslilly5834 do you understand how technology works? Once new technology is found, you don't have to relearn it. Of course Tesla is going to have easier time creating a bot from the 90s, when the research was already done 30 years ago. It's not a toddler learning to walk, it's a machine based on technology we've had for a long time.
I work in the ai industry and “robots do jobs people don’t want” is just a diplomatic way of saying you’re not needed anymore cause robots can do your job more efficiently for next to nothing. The number of people sacked due to automation is staggering
that's just one thing, everything in Boston turned out to be scripted and scripted and yes, they went bankrupt and the Koreans bought them out...Hyundai Carl!
@@mrb2349 yes. We can see that in FSD that is working completely autonomously from 2020 like Elon promised right? And it doesn't say in manual about all the exceptions that can happen right?
that's just one thing, everything in Boston turned out to be scripted and scripted and yes, they went bankrupt and the Koreans bought them out...Hyundai Carl!
The Tesla is already the better robot. Go watch them in person, if they will let you. What happens constantly with the Boston robot is that it stumbles, falls, fails, etc., far, far more often than it succeeds.
BD retired Atlas because it has no practicality, inefficient, prohibitively expensive, dangerous with hydraulic system. So much for like life movement. Tesla said no thanks to that relic. Good thing BD took inspiration and developing Atlas 2.0.
@@nguyep4lol if you think Boston Dynamics highly articulate and mobile Atlas 2 took inspiration from the geriatric stiff walker known as Optimus you’re very delusional. The moment Boston Dynamics makes one video with Atlas 2 they’ll still be well over a decade ahead of Tesla.
@@nguyep4 atlas 1 are just for exp. Just like spot where made. Thry had several bot before they release a better version. Atlas 2 is not something they just did because others took that diff route. Atlas 2 has been in the making way ahead of other company. I wouldnt be surprise if they have several other bots that they are not sharing publicly.
@@nguyep4😂 you probably don't even mix water with Kool-aid, you just dump the powder right down your throat and ask for more huh. Don't forget to provide another bonus to Elon that exceeds annual profits by 3x. 🤭👍
I’m not convinced that the Tesla Bot is a serious project. Never seen it do anything physically interesting. Pretty sure it’s just a PR thing for Elon. Atlas is ALWAYS impressive, though. They really thought outside the box with this one already.
They actualy made robot. Its just its about 20 years behind Bostons Dynamic. It works only in pre-programed enviroment with no obstacles. It can only move in pre-programed path and needs markers in the enviroment to know where it is and where it should move. And even with all this, it still needs multiple attempts to move from point A to point B while holding box without some kind of problems. You can find their video on youtube. And that is probably all that we ever will see from Musk on robotics. He made big promises, delivered pre-pre-pre-prealpha product and now he will pretend as if he never bragged about his venture into robotics
@@iglidor while I agree that Tesla robot is significantly behind in terms of hardware. 20 years is a massive overstatement. Also the claim that the Tesla robot only operates in preprogrammed ways couldn't be farther from true. The whole point of the robot was, that Tesla is extraordinarily good at mapping the outside world into a model, a computer (car or robot) can use. I wouldn't be surprised if the software of the Tesla bot is actually superior to Boston dynamics software. You wouldn't notice that fact though because everything hardware related is so far behind.
@@brockmann4815 Have you watched their promotional videos where they showcased their robot? There are distinct markers placed in places where they want their robot to go. Because their robot was unable to properly map outside world. It is extraordinarily BAD at mapping outside world. And that is in THEIR promotional video where they were showing only what they wanted and where they covered as much problems as they could. And still failed since they were kinda lazy about it and it was obvious that their video is spliced from multiple attempts because they could not get even a single clean run of their test.
@@iglidor Hate on them all you want, but something tells me you have no idea about image recognition. I'm actually studying in that field and worked with image recognition. Markers are used to train the robots neural network object recognition. In their showcase from 11 months ago you actually see the international representation which doesn't use markers at all. It places dots where the area is obstructed and that's essentially how such a representation is implemented. The Tesla fsd does it the same way although in a more complex manner. So now add a few years of tweaking that software to the requirements and sensors of the robot and you get a fairly capable model. The first step is representing physical space, the second step is object recognition, and the third step is efficiency and capability of movement. The Tesla bot is progressing at a far faster pace than Boston dynamics. It'll probably take years before they have a compatible model, but certainly not decades. You seem to be biased on the topic too. You seem to hate Tesla itself and therefore insist that Tesla produces 'lazily done' videos where they can't get a single perfect take. That's just how musks companies operate. Hardware and failure rich testing improves the rate of improvement.
Then again, I say it's perfectly fine if they take the jobs that often end in human injury or death. Underwater welding is the first that comes to mind. What I DON'T want to see is robots taking jobs for human creativity (AI artists/writers in Hollywood.)
@@Tonabillity No. As I said, it's inevitable that technology replaces jobs. But we shouldn't deny that is what's happening and it's a problem. Some people try to play it down and deny the negative impact. I'm just suggesting we do things mindfully with eyes wide open.
I kind of wonder if/when VR tech will start being used in conjunction with these kinds of bots. I could see it for situations where the human body is just very well suited to the situation, but the environment itself is hostile to the human body, so they send in one of these Atlases instead but control it with a VR headset and controllers
The Dawn Cafe in Japan employs around 70 disabled or disadvantaged people VR/RC operate robotic servers from the comfort and safety of their own homes,allowing a partially paralyzed young man to financially contribute to his family's income while enjoying the socialization of his job. A lady who lives in Australia also waits tables there. No commuting required! Check out the beta stage "telepresence operation of an Asha robot" video posted a few year's ago on RU-vid. All 3 of Hanson's Sophia Asha and Grace The Humanlike Nursing Assistant are about to be outmoded by their new next level advanced walking version mass production model "Project RIA" in a collaboration between Hanson Robotics and Machani Robotics. The multipurpose Beomni robot has already been signed up for a record 1,000 unit contract, with an option for 10,000 more within 5 year's. Several other high end research teams also have similar abilities as well as autonomous programming and a combination of both operational modes.
The tesla bot will be a mass produced with die cast and plastic injected parts. The precision parts in the evaluators will be what drives the cost. If they can punch these out cheaply, the robot should be affordable. The Boston Dynamics robot doesn't look like it's designed for affordability.
Don’t be fools. Eventually robots will take jobs. Why would a company not invest in robots that need no sleep, no insurance no food essentially no comfortable heat possibly. It’s a matter of time. The question will be how will displaced workers be taken care of.
When U.S. Steel made building bridges cheap, safe, and reliable people said the same thing about the ten of thousands of ferries. Before that, it was the loss of all the blacksmith jobs in the liberty stables.
@@Griffix96 this is completely different. This isn't the horse vs Ford Model T. this is a tired horse vs a Telsa with max range and autopilot. The push is for "efficiency" and the get better argument does not work across the board for all humans. The guy doing construction won't be the guy programming the bots later.
B.D. has been doing this for decades already. Literally. But its mainly programming/directed movement more focused on the mechanics. Tesla is newish to the game and is starting with self guided robotics. Honestly if you put them together youd have a halfway decent terminator
New Battlebots FullSize! In the Red corner, Tesla's Optimus with a sword, VS Atlas by Boston Dynamics, with just a shield, in the blue corneeeeeerrrrr.
The current rate of innovation does not predict future rates, I expect it to look like a logistical curve in the long term. The question remains where on the curve is each company and where will they end up on the curve in 5 or 10 years.
Boston Dynamics has/had a DARPA contract so that funding will push it past most the competitors. Even though the Mule is not the same as these humanoid style robots, the funding definitely helped the conpany as a whole.
If any company can save money on labor, they will. As long as maintenance isn't as expensive as a human employee the robot will win hands down. For job security I recommend to be a technician, least til the robots start repairing themselves.
I could see the Atlas models in an industrial setting. But I can also tell which of the two is likely to be the consumer model that sells 100x more units. Especially if only one of them gets access to AI training on custom-made AI chip arrays.
Boston dynamics have been doing this for over 15 years. Tesla have been doing this for 2.5 years and look at the strides they have made since Optimus v1. The way Tesla work, they will be light years ahead soon.
The Tesla bot was the best when it was first introduced it broke down dancing! I was blown away it could move so fluidly, but then they slowed Optimus down! Why? The bot is now a snails pace and it can't dance!
As long as the job people don't like is still a job that people can do, a robot performing the task will always take someone's job from them. I don't like this at all, lord know we have enough unemployed. But sadly this is nothing new, we been getting rid of or streamlining jobs with technological advancement since forever.
the part most people skip over is that mass producing robots is going to take a long time. I think by the time they can actually threatened jobs or make a large scale impact most of us will be long retired. Just make sure your kids can repair robots : p
Seriously though, seeing how each robot tackles human-like tasks makes you wonder about the future of automation. Will they take our jobs, or will they become our quirky co-workers? Either way, I'm just glad I won't be stuck in an office with either of them! What do you all think? Are we heading towards a harmonious robot revolution, or is this just the beginning of their uprising?
I'm all for robots taking our jobs, I mean there's a massive labor shortage and has been for years. But can we all agree these things need a speed limit? If they can go faster than 5mph that's an issue
The Tesla bot may take away our factory jobs, but the follow-on ATLAS has the potential to take away more general and broader categories of jobs done by humans such as combat soldiers clearing muddy trenches.
I think you don't understand much about robotics, or understand that Tesla has made something very rudimentary. Tesla is in major financial problems right now, and their bot is basically a grift on stock price. It's a car company.
as an enterpreneur, i choose to have optimus, with ai learning capabilities, ai software really change the way i see a company works daily, i can reduce the number of people i should hire to max my company profit, by eliminating menial jobs and position, has reduce my company operating cost, such as labour cost, insurance package, allowances, to renting workspace since we can work from anywhere. now i can't be bothered with labour laws, etc. especially the cost (time, the effort to cultivate human resources, etc that can't be determined by money but have to be converted to currency to exactly know the cost for each project) for everything, dinamyc of human relation between co-worker, lawsuit that may occur because of false accusation that can jeopardize my company reputation, etc.
Teslabot will likely be way too expensive for real world use for longer than either Atlas or Digit. Though that's not to say those other units aren't expensive either. One electric Atlas unit is around $45,000 a piece
AI in general was pitched as augmenting people .. well .. until bosses started firing tens of thousands of workers to replace them with ai tools customers would interact with .. same will apply, it’ll come down to replacing people if it’s cost effective enough, it’s never for the workers benefit
Everyone forgets that humans weren't born on earth to work labor and serve the rich. So I hope robots take all the jobs. So we can be free and have actual time to live our lives
How it plays out is a bit like crossing demolition man, blade runner and terminator with a side of total recall if Elon colonises mars before the terminator part that is. Interesting times
It’s happening y’all. Both Robotics designs and AI are making tremendous strides in development. It’s only a matter of time until they bring these two things together, then the Future begins… we may need some John connors 😬
If a robot can do a job, don’t need to be paid monthly, don’t takes holiday, don’t organise themselves in union, end up being the cheapest labour ever, and eventually with higher productivity, which CEO will say no to that we have to keep human in the loop?
The key difference between Boston Dynamics' Atlas and Tesla's Optimus is that Atlas is pre-programed, whereas Optimus is powered by artificial intelligence. The former focuses on robotics while the latter focuses on AI. Tesla said that Optimus was designed to perform tasks that were "unsafe, repetitive, and boring"; Boston Dynamics said Atlas was designed to perform tasks that were "dull, dirty and dangerous". There is virtually no difference in the two companies' descriptions of Atlas and Optimus. But given that Atlas is pre-programed, I think Atlas is more suitable to perform high-speed, high-precision, and repetitive tasks, while Optimus is more suitable for versatile environment where some sort of judgement is required, because it is powered by AI. Other things alone, there are two areas where Optimus have clear advantages over Atlas: (1) Affordability. Atlas costs up to $150,000 USD, whereas Elon Musk plans to mass produce Optimus for under $20,000 each. (2) Atlas is powered by high-powered hydraulics, whereas Optimus benefits from Tesla's existing car technology and battery efficiency, allowing it to operate for extended period of time. Great content. Thanks for sharing!
An update to my own comment: I just found out that Boston Dynamics released a fully electric robot in April of this year. The hydraulic version of Atlas had been retired.
