@@garg414 Well, I'm not accusing 1X of faking these videos. I'm just saying how it looks. But if one were going to fake it, you would conceal the guy's hands at the end of the forearm of the robot suit and have a robot claw or hand sticking out from the end. The hands of the guy in the suit would clutch a mechanical grip that would cause the robot claw or hand to open and close. Of course, this would make the robot's forearms inordinately long and thick at the wrists. You know, just like the ones in this video (0:52). If Neo allowed its arms to hang straight down, the tips of its fingers would be almost at its knees. Compare Optimus with Neo (4:57). But I am not saying they are faking it. I'm just describing what I see.
for me optimus is better...last video optimus was actually moving while standing making it appear as if it is constantly making adjustments to keep balance like a real person...cant wait for next update of optimus
My biggest question is security. What processes/policies do companies have in place to ensure that the video data sent back to the centers is secure from employees and hackers....
All of their hardware? Everything in NEO is made in house at 1X production facility, planed and designed for mass production from day 1. They have many patents on groundbreaking vital components and engines, shy they can build their NEO half the weight of Optimus, and 1/4 the weight of other competitors… This is one of the reasons NEO is safer in homes. If a robot falls over a child you want it to be as lightweight and soft as possible.
I get the impression that, aside from Tesla, most of the humanoid robot companies make only humanoid robots. This means that while they are still in the early stages of scaling up production, they need to build/buy batteries, sensors and AI computing chips in relatively small batches. In contrast, Tesla is already building/buying batteries, sensors and AI computing chips for millions of cars, so Tesla has economies of sale in its building/buying of those components. This can help give Tesla a cost advantage, at least until its competitors scale up to producing millions of robots.
At 6:30 I doubt that over doubling of weight of Optimus is because of its outer shell. That does not make sense to me. I figure it is most likely because of the massive 2.3 KW batteries lasting over 6 hours versus the other just lasting 2 hours so the battery is probably less than half the size/weight. Also, Tesla's Optimus probably has much higher load door capacity in its motors which will make them much heavier. Optimus is more of an industrial manufacturing robot versus the other is just a household kind.
Well, heck, if the robot is being controlled by telepresence, then of course it (or rather, the telepresence operator) can respond to voice "commands" and both initiate and respond to gestures as well as identify and manipulate objects (like in a refrigerator) or do most anything a person can do. I had recently seen the promotional video with the girl and backpack and been impressed. But if that was telepresence, it would be normal and not "autonomous" or particularly robotic at all. For example, the Canadian Arm on the ISS is often referred as a "robotic arm" but it really is a "Waldo" or a prosthetic device being controlled remotely or directly connected by a person. Optimus can also be controlled by telepresence and that aids training. However, that is not the same as when it is walking around doing things/achieving goals autonomously based on sensor inputs and inference engine recognition, planning, and agency control. I will say, though, that the hardware platform for Neo (using tendons/"strands", etc., and being a bit lighter) and probably the compute platform (providing "autonomic" nervous system-like control of balance and movement) do both seem pretty good. But, the inference engine? Maybe not.
Not so much talked about, but some former top guys from Tesla and BMW have moved to 1X, and are now leading the search for finding locations for production and scaling. There is a lot production and design experience at 1X now. I think people that dissmiss 1X don’t have clue about the company.
@@Cleanerwatt Fair enough, I thought it was an explicit denial to use snippets. They do great videos, I was just gonna "boycott" this one in particular Great comparison video btw, thanks!
In terms of function wheels rather than feet strikes me as being far more efficient and effective dependent on purpose. Working in and around a factory or house on simple functions like carry, transport, respond to environmental conditions such security, closing doors, cleaning, filming or specialist tasks inside toxic or hostile environments wheels are a great option.
I think we should keep robots like this forever because if they get too advanced it will be too much especially if we have robots everywhere and they become more advanced with their movements
You should do a new review of the NEO after watching the video of Nick Giovanni in a cooking competition against NEO. I was very impressed. I'd like to see your reaction.
Thanks for informative video! I think physical safety will be an important aspect going forward. Most humanoids are heavy and uses motors and gears, and are thus not safe for human environments. (The momentum in movement is to strong, and they can crush people if falling over) Neo is designed for interaction with humans, and can hopefully be applied in for instance the health sector. But it might not be the best fit for industrial use.
It's pretty obvious that the 1X Neo is screwed to the floor so it doesn't fall over when picking up the bag. Lol. I'm glad more companies are getting in the game and while I definitely think they are on to something going the "tendon" route you can tell at this point they have absolutely no precision control in that tendon/actuator movement. They have almost no software at this point either compared to the "real" AI companies...so ya.. kinda ridiculous to put it in the same league as Optimus or Figure 2.
I think they understood they were on the wrong path, and now Tesla copy 1X. It will be interesting to see if Optimus 3 will be more like NEO. Anyway, 1X is already working on NEO 2 gen, which the CEO says will be a lot better.
We are a long way from knowing which one is better. Tesla hasn't shown what Optimus can do, and 1x is more pretend than actual performance. Of these two, I'm more convinced that Optimus is the real thing than that 1x is. But I think anyone can build a humanoid body. Giving it a brain, though,...
Any comparison at the moment is on shaky ground at best. Why? Because we don't actually have a product. All we have are various prototypes with various claims as to their respective actual future product. And I have come to be EXTREMELY skeptical as to the validity of some of these smaller startups. I think they are vastly underestimating what it ACTUALLY takes to do what they claim they want to do. Ask Elon what it ACTUALLY took to ramp the model 3 or what it ACTUALLY took to land the Falcon 9. All we have in regards to these other startups are highlight reels of clumsy robots with dubious autonomy, no concrete path to mass manufacturing, no clear path to mass data flywheel, no clear path to an exaflop-scale training cluster, no supply chain and no information about the projected unit costs or how they will survive the ramp. IMO Tesla is the only company that has the credibility to do all that is required to get this ACTUALLY to mass scale. (Some chinese companies might as well, that's all a bit unclear.)
What matters is the REVENUES from the Bot, the tech - sure but not so much - Tesla will be SUPERIOR - so....what matters is the REVENUES, please cover!
Is it legal to take frame by frame screenshots of a video you don't have permission to show and flip through them at 24 fps to give the impression of video?
"Neo has demonstrated it can respond to voice commands but it may have been a human responding via tele-operation." What?? NO. If we're not sure whether the robot was acting autonomously or was just a remotely operated puppet, then Neo has NOT DEMONSTRATED any such capability. By definition. That's what demonstrated capability means - proving that it can actually do the thing - as opposed to a simulation, mockup, scam, fraud, computer render, powerpoint slide.
I cannot see the upside in these to even get these sold. For example why would people or companies invest in buying these except to waste money for a hobby. At this stage they have way to far of a gap to even be relatively useful. It would one thing if i could leave it at the house and come home to chores being completed and dinner cooked. I am not sure of the type of menial tasks that would justify the purchases i guess is my point.
The flood of humanoid robots being revealed lately is exciting, but let's be honest, they are just prototypes operating in benign, controlled environments. There is only one robot that can handle real-world environments: robot cars from Tesla. And even they have not achieved full autonomy yet. Achieving real-world autonomy is not so much about the electro-mechanical design of a robot. That's the easy part, as demonstrated by numerous robotic startups, 1X included. The hard part is creating the intelligence needed to understand and interact with the real world. Building a neural model of the world with all its intricacies, will require enormous investments in data collection and training computational resources. Without that, scaling up production for economic advantage will be impossible. Comparing 1X NEO with Tesla Optimus while ignoring this fundamental aspect of what will make humanoid robots practical, whether in the factory, retail establishments, or the home makes any comparison meaningless.
Fair use is a thing. I really dislike demos like this... i want to see REAL WORLD uses for a robot... not giving a hug, or doing something my smartphone does faster/better... idk
So this is just another disingenuous Tesla fanboy video. Things deliberately omitted in this video: 1. Neo can deadlift 70kgs 2. Neo was using gpt4o for voice and in future will use a custom voice model 3. The body of Neo is designed to be safe and usable in homes on Day One unlike Optimus which is not human safe. Also, unlike Tesla 1x already sells commercial robots and has plans in place for at scale production.
@@danielcpt3819 c'mon man. Learn about robotics rather than going the crazy fanboy route. I'll rather be civil instead and try to educate. Current materials and actuators used for Optimus are heavy and too powerful like all older robotic architecture. Thereby it is dangerous to use around humans. The current Optimus model is designed to work in factories without any humans in the proximity. Neo on the other hand is made from softer and lighter materials with additional soft coating to be able to directly integrate in the household. Hope this makes you feel better.
@@vicdelta31415 Saying Optimus isn't "human-safe" is inaccurate though as it is designed to not be dangerous to the human companion it will be working alongside and taking instruction from.
Neo looks like it's a guy in a robot suit.
how would a guy in a robot suit make 90% of his thumb disappear
@@garg414 Well, I'm not accusing 1X of faking these videos. I'm just saying how it looks. But if one were going to fake it, you would conceal the guy's hands at the end of the forearm of the robot suit and have a robot claw or hand sticking out from the end. The hands of the guy in the suit would clutch a mechanical grip that would cause the robot claw or hand to open and close. Of course, this would make the robot's forearms inordinately long and thick at the wrists. You know, just like the ones in this video (0:52). If Neo allowed its arms to hang straight down, the tips of its fingers would be almost at its knees. Compare Optimus with Neo (4:57). But I am not saying they are faking it. I'm just describing what I see.
The "Fisker" of humanoid robots
barely agree.
for me optimus is better...last video optimus was actually moving while standing making it appear as if it is constantly making adjustments to keep balance like a real person...cant wait for next update of optimus
Using tesla self driving O should be able to navigate "real world" much earlier.
My biggest question is security. What processes/policies do companies have in place to ensure that the video data sent back to the centers is secure from employees and hackers....
Same as "alexa" not
Surely there’s a cost savings to be had by vertically integrating the hardware for Tesla. It seems like NEO is outsourcing all of their hardware.
All of their hardware? Everything in NEO is made in house at 1X production facility, planed and designed for mass production from day 1. They have many patents on groundbreaking vital components and engines, shy they can build their NEO half the weight of Optimus, and 1/4 the weight of other competitors… This is one of the reasons NEO is safer in homes. If a robot falls over a child you want it to be as lightweight and soft as possible.
I get the impression that, aside from Tesla, most of the humanoid robot companies make only humanoid robots. This means that while they are still in the early stages of scaling up production, they need to build/buy batteries, sensors and AI computing chips in relatively small batches. In contrast, Tesla is already building/buying batteries, sensors and AI computing chips for millions of cars, so Tesla has economies of sale in its building/buying of those components. This can help give Tesla a cost advantage, at least until its competitors scale up to producing millions of robots.
I think we should have humanoid robot wars
At 6:30 I doubt that over doubling of weight of Optimus is because of its outer shell. That does not make sense to me. I figure it is most likely because of the massive 2.3 KW batteries lasting over 6 hours versus the other just lasting 2 hours so the battery is probably less than half the size/weight. Also, Tesla's Optimus probably has much higher load door capacity in its motors which will make them much heavier. Optimus is more of an industrial manufacturing robot versus the other is just a household kind.
Well, heck, if the robot is being controlled by telepresence, then of course it (or rather, the telepresence operator) can respond to voice "commands" and both initiate and respond to gestures as well as identify and manipulate objects (like in a refrigerator) or do most anything a person can do. I had recently seen the promotional video with the girl and backpack and been impressed. But if that was telepresence, it would be normal and not "autonomous" or particularly robotic at all. For example, the Canadian Arm on the ISS is often referred as a "robotic arm" but it really is a "Waldo" or a prosthetic device being controlled remotely or directly connected by a person. Optimus can also be controlled by telepresence and that aids training. However, that is not the same as when it is walking around doing things/achieving goals autonomously based on sensor inputs and inference engine recognition, planning, and agency control.
I will say, though, that the hardware platform for Neo (using tendons/"strands", etc., and being a bit lighter) and probably the compute platform (providing "autonomic" nervous system-like control of balance and movement) do both seem pretty good. But, the inference engine? Maybe not.
thanks for info
Not so much talked about, but some former top guys from Tesla and BMW have moved to 1X, and are now leading the search for finding locations for production and scaling. There is a lot production and design experience at 1X now. I think people that dissmiss 1X don’t have clue about the company.
Kinda odd S3 didn't want you to use snippets from their videos... For sure I'm not gonna watch their video because of that
I asked, but never heard back from them... I would still watch their videos.
@@Cleanerwatt Fair enough, I thought it was an explicit denial to use snippets. They do great videos, I was just gonna "boycott" this one in particular
Great comparison video btw, thanks!
In terms of function wheels rather than feet strikes me as being far more efficient and effective dependent on purpose. Working in and around a factory or house on simple functions like carry, transport, respond to environmental conditions such security, closing doors, cleaning, filming or specialist tasks inside toxic or hostile environments wheels are a great option.
I think we should keep robots like this forever because if they get too advanced it will be too much especially if we have robots everywhere and they become more advanced with their movements
That's an impossibility. Companies are in competition for the dollar and competition drives innovation
Come on, we all want to know when we're literally going to see these two bots fight. Humanoid Battlebots! (Actually, that's a pretty scary thought!)
You should do a new review of the NEO after watching the video of Nick Giovanni in a cooking competition against NEO. I was very impressed. I'd like to see your reaction.
Thanks for informative video!
I think physical safety will be an important aspect going forward.
Most humanoids are heavy and uses motors and gears, and are thus not safe for human environments. (The momentum in movement is to strong, and they can crush people if falling over)
Neo is designed for interaction with humans, and can hopefully be applied in for instance the health sector.
But it might not be the best fit for industrial use.
FYI John, It's "Across", not "Acrossed" or "Acrosst".
It's pretty obvious that the 1X Neo is screwed to the floor so it doesn't fall over when picking up the bag. Lol. I'm glad more companies are getting in the game and while I definitely think they are on to something going the "tendon" route you can tell at this point they have absolutely no precision control in that tendon/actuator movement. They have almost no software at this point either compared to the "real" AI companies...so ya.. kinda ridiculous to put it in the same league as Optimus or Figure 2.
I don't know about being screwed to the floor. I didn't see anything that would suggest that. But I agree. 1X is definitely putting up a heavy front.
Tesla/Optimus is going to use tendons for hand control too. I wonder how they thought of that?
I think they understood they were on the wrong path, and now Tesla copy 1X. It will be interesting to see if Optimus 3 will be more like NEO. Anyway, 1X is already working on NEO 2 gen, which the CEO says will be a lot better.
We are a long way from knowing which one is better.
Tesla hasn't shown what Optimus can do, and 1x is more pretend than actual performance.
Of these two, I'm more convinced that Optimus is the real thing than that 1x is.
But I think anyone can build a humanoid body. Giving it a brain, though,...
1X says NEO's promised walking speed will be achieved by the end of the year.
Any comparison at the moment is on shaky ground at best. Why? Because we don't actually have a product. All we have are various prototypes with various claims as to their respective actual future product. And I have come to be EXTREMELY skeptical as to the validity of some of these smaller startups. I think they are vastly underestimating what it ACTUALLY takes to do what they claim they want to do. Ask Elon what it ACTUALLY took to ramp the model 3 or what it ACTUALLY took to land the Falcon 9. All we have in regards to these other startups are highlight reels of clumsy robots with dubious autonomy, no concrete path to mass manufacturing, no clear path to mass data flywheel, no clear path to an exaflop-scale training cluster, no supply chain and no information about the projected unit costs or how they will survive the ramp. IMO Tesla is the only company that has the credibility to do all that is required to get this ACTUALLY to mass scale. (Some chinese companies might as well, that's all a bit unclear.)
……… yeah what dey said. !!!!!!
What matters is the REVENUES from the Bot, the tech - sure but not so much - Tesla will be SUPERIOR - so....what matters is the REVENUES, please cover!
Is it legal to take frame by frame screenshots of a video you don't have permission to show and flip through them at 24 fps to give the impression of video?
That's just a video
"Neo has demonstrated it can respond to voice commands but it may have been a human responding via tele-operation."
What?? NO.
If we're not sure whether the robot was acting autonomously or was just a remotely operated puppet, then Neo has NOT DEMONSTRATED any such capability. By definition. That's what demonstrated capability means - proving that it can actually do the thing - as opposed to a simulation, mockup, scam, fraud, computer render, powerpoint slide.
Competition is good. That market is too big, a lot of space for many winners.
“None the less”
the robotics and capabilities of Neo look quite disappointing at this stage. He didn't do anything in those 48 hours much more than making coffee.
Leon always claiming to have the best, be the best. I'm reminded of the bankrupt hyperloop, the crater under the OLM, the flame coaches. Pass.
I cannot see the upside in these to even get these sold. For example why would people or companies invest in buying these except to waste money for a hobby. At this stage they have way to far of a gap to even be relatively useful. It would one thing if i could leave it at the house and come home to chores being completed and dinner cooked. I am not sure of the type of menial tasks that would justify the purchases i guess is my point.
What i want to see is it making a sandwich... and then cleaning up.
The flood of humanoid robots being revealed lately is exciting, but let's be honest, they are just prototypes operating in benign, controlled environments. There is only one robot that can handle real-world environments: robot cars from Tesla. And even they have not achieved full autonomy yet. Achieving real-world autonomy is not so much about the electro-mechanical design of a robot. That's the easy part, as demonstrated by numerous robotic startups, 1X included. The hard part is creating the intelligence needed to understand and interact with the real world. Building a neural model of the world with all its intricacies, will require enormous investments in data collection and training computational resources. Without that, scaling up production for economic advantage will be impossible. Comparing 1X NEO with Tesla Optimus while ignoring this fundamental aspect of what will make humanoid robots practical, whether in the factory, retail establishments, or the home makes any comparison meaningless.
I think neo is better
Fair use is a thing.
I really dislike demos like this... i want to see REAL WORLD uses for a robot... not giving a hug, or doing something my smartphone does faster/better... idk
Aren’t Tesla Optimus videos in 1.25x speed?
tesla should buy this company
I don’t think Samsung, Open AI, Nvidia, Tigerglobal, founder and other investors are interested in selling to Tesla, ha ha
Humanoid robots should be limited to walking no faster than 4 MPH. We need to know we can run away from it, if necessary.
That's crazy! What if you're a child, or someone in a walker?
@@BrianBellia Then you're SOL!
fake
So this is just another disingenuous Tesla fanboy video.
Things deliberately omitted in this video:
1. Neo can deadlift 70kgs
2. Neo was using gpt4o for voice and in future will use a custom voice model
3. The body of Neo is designed to be safe and usable in homes on Day One unlike Optimus which is not human safe.
Also, unlike Tesla 1x already sells commercial robots and has plans in place for at scale production.
I suppose Optimus has been designed to kill humans has it? Get a life.
@@danielcpt3819 c'mon man. Learn about robotics rather than going the crazy fanboy route. I'll rather be civil instead and try to educate. Current materials and actuators used for Optimus are heavy and too powerful like all older robotic architecture. Thereby it is dangerous to use around humans. The current Optimus model is designed to work in factories without any humans in the proximity. Neo on the other hand is made from softer and lighter materials with additional soft coating to be able to directly integrate in the household. Hope this makes you feel better.
@@vicdelta31415 Saying Optimus isn't "human-safe" is inaccurate though as it is designed to not be dangerous to the human companion it will be working alongside and taking instruction from.
@@vicdelta31415you're clearly an anti Tesla troll.
This is such a silly comment.
Where does one even begin with how ridiculous this is?
🥳❤💢💥💫👍