Best Humanoid Robot in 2024: Tesla Optimus vs Figure 2 | Head to Head Comparison

[@ricardoveras3433]

In the future these companies should actually have competitions with their humanoid bots like a sewing challenge and other complex tasks. Would be fun and incentivize innovation

[@zaphodbeeblebrox2817]

DARPA did that a decade ago.

[@ricardoveras3433]

@@zaphodbeeblebrox2817 with technical humanoids?

[@joechughtai3155]

I think you touched on the key point right at the end. COST, if Figure can build a 5% better bot but it costs 50% more than Optimus it's not gonna matter to most people.

[@klauszinser]

I don't think figure has advantages.

[@jay48252]

@@klauszinserI disagree figure has a much better software running it than Optimus in my opinion.

[@jmattoxriskpro]

@@jay48252what do you know about the software running Figure?

[@mannygee005]

Tesla has the lead in manufacturing. If there are 20 ingredients to mass manufacturing then Tesla has a 20x lead.

[@alphaomega1969]

No, china is

[@alphaomega1969]

Tesla is full of BYD parts also china is the only country to use 95% full automation in production facility for maximum profit which Tesla is using. Tesla, spaceX, Starlink, Solar City, Neuralink, The Boring Company and Optimus Robot are all Chinese startup investment to boost US economy due to trade deficit. Elon Musk would be no one without china. lol

[@alphaomega1969]

Both Figur and Optmus bot are Chinese outsourced tech, in a nutshel

[@TroySavary]

Boston Dynamics is owned by Hyundai, who manufactures far more than Tesla.

[@charlescole-p9v]

Tesla will have the edge with it's MASSIVE inference compute. And the best engineers on the planet.

[@feet_first]

Best engineers on the planet?🤣😂😆🤣

[@TroySavary]

The falling apart Cybertruck makes me question that "best engineer" claim.

[@silaskelly604]

As far as I know, Optimus uses the same inference computer that the Tesla cars use and they understand text commands really well.

[@jochenvonbastianeller6865]

And they have the compute power and the data. And " prototipes are easy production is hard "

[@johannesdolch]

And they are the sole leader in actual manufacturing. Anybody who tries to compete in EVs is now going bankrupt.

[@TroySavary]

​@@johannesdolchNobody is going bankrupt because of Tesla. Delusional fanboy nonsense.

[@andrew390]

Great video! Keep the regular Humanoid Robotics update videos coming please.

[@olyalphy]

I think Tesla will have an excellent utilization rate for Optimus on their factories. This is thanks to their in-house factory orchestration software, managing materials, labor, downtime, tracking a digital twin of assets in real time, it will truly be a dreadnaught factory with hardly any humans in sight!

[@momohnyaley7069]

I've never seen a robot without flexed knee in it's default standing position. That's shows it's difficult to make them stand fully upright. So we should thank God that we were created to stand upright. Don't take it for granted.

[@jamessullivan7964]

Good point now it looks like we’re playing god

[@bobsalita3417]

Great guest, Dr. Scott Walter. Hadn't heard of him before.

[@Cleanerwatt]

He is knowledgeable and definitely worth following!

[@TheFutureThoughtExchange]

Degrees of freedom (DoF) in robotics typically refer to the independent movements a system can make. When it comes to a robotic hand, having 22 DoF means there are 22 independent movements, usually corresponding to joint motions or combinations of joint motions. Let’s break down a possible configuration of 22 DoF for a robotic hand, approximating the degrees of freedom in a human hand: 1. Thumb (4 DoF): • Carpometacarpal (CMC) Joint: • 2 DoF: Flexion/Extension (move thumb across palm) and Abduction/Adduction (move thumb away/toward the palm). • Metacarpophalangeal (MCP) Joint: • 1 DoF: Flexion/Extension (bend thumb at the knuckle). • Interphalangeal (IP) Joint: • 1 DoF: Flexion/Extension (bend tip of thumb). 2. Index Finger (4 DoF): • Metacarpophalangeal (MCP) Joint: • 2 DoF: Flexion/Extension and Abduction/Adduction (bend at the knuckle and move side-to-side). • Proximal Interphalangeal (PIP) Joint: • 1 DoF: Flexion/Extension (bend the middle joint). • Distal Interphalangeal (DIP) Joint: • 1 DoF: Flexion/Extension (bend the tip joint). 3. Middle Finger (4 DoF): • Same as the index finger: 4 DoF (MCP, PIP, and DIP joints). 4. Ring Finger (4 DoF): • Same as the index finger: 4 DoF (MCP, PIP, and DIP joints). 5. Little Finger (4 DoF): • Same as the index finger: 4 DoF (MCP, PIP, and DIP joints). 6. Wrist (2 DoF): • Flexion/Extension (bending the wrist forward and back). • Ulnar/Radial Deviation (moving wrist side to side). Total: 22 Degrees of Freedom. This configuration approximates the anatomy of a human hand. In robotic design, additional degrees of freedom may sometimes be added for specialized tasks or to give a hand more dexterity. Some hands might also allocate fewer DoF to the wrist and more to the fingers, depending on the design goals.

[@Plazman]

Don't forget Unitree. On the affordability front, they're CRUSHING it.

[@bru512]

The real key question How good are the hands at handling items? That is the key metric. The rest is secondary

[@leonard9688]

I think optimus will be at least 5 years ahead of fig 2 in regards to mass production and final cost. I think that will be the big difference holding back figure 2.

[@eriona1007]

Have you just ever thought that Tesla is already focusing on the cost instead of just some 1%～5% function improvement ?

[@Urgelt]

My question is, does Figure 2 build a virtual environment to reason about the actual environment it is in? Probably not if LLM inference is the onboard processing concept. FSD does that. I'm pretty sure Optimus will, too.

[@BillSmith-fx7xx]

HINT: Y'all engineers, saw the hands had to be done like real hands, no shortcuts. I hope you're working overtime on the feet ! :- )

[@jonathanlivingston7358]

How could you not talk about FSD? That is the biggest difference between Optimus and any other driving or robotic companies. Tesla is an expert in real life video training and has dojo

[@TroySavary]

The FSD that is under investigation for killing people? That FSD?

[@verynice5574]

Nevermind that a human could perform the battery placement task 10x as fast in a real world scenario without supervision.

[@mich8411]

But can they do it for 24 hours straight, Nop not to mention the speed of those Robots will also 10X from now while they will also be getting way cheaper so CHECKMATE

[@TroySavary]

​@@mich8411A purpose built sorting machine can do it all day, 100+ times faster, and for less money than Optimus. Industrial robots have existed for decades and are far better suited for the task.

[@jmattoxriskpro]

This is basically a toddler level experience robot. To judge it as an idea on this current performance is ill informed

[@TroySavary]

@@jmattoxriskpro Judging Optimus on its current performance is wiser than judging it based on what a compulsive liar claims it will do.

[@KB-uw6ne]

Must you be connected to the cloud by WiFi, or can you use Ethernet cable? And what are the decibels levels of servo motors, and actuators?

[@jmattoxriskpro]

Quiet

[@mannygee005]

Good job Dr Scott Walter. Nice resource. Edit - what might be missing as the actuators. Tesla are designing their own actuators for mass production. And... in innovation and iteration, there are none faster than Tesla/SpaceX/xAI. Figure AI will have to go head to head vs the Tesla style organization, sounds like a good time. It will be a challenge for any company.

[@User.70793]

how long until 100% safe level 5 autonomous driving! and what level of autonomous driving are we at now? Has Tesla reached an advanced level 3? or are these shades of a level 4? (I mean the beginning)

[@johnsonjjohnson100]

I think Figure has the edge right now and are already being used in BMW factories Plus they partnered with an automotive company to mass produce their bots Thus they have come farther, faster and have partnered with the biggest AI expertise in the world

[@mich8411]

The moment you mentioned Partner, then Figure advantages fades, Figure have to Partner in many things and will be extremely Expensive, Tesla Optimus will be 40% cheaper if not 30% because of their In house expertise

[@gemin0i]

Figure the GOAT

[@luke_fixed5266]

Robots taking recharge breaks? Really? Easier to program in tasks in a localized domain that allow the robot to tether and charge as it works.

[@ModernCountry365]

Not when it's working around your house.

[@johnsonjjohnson100]

Wireless induction charging Robot walks up to the charging station and it starts charging through its feet Tesla bought a wireless charging company somtime back

[@ModernCountry365]

@@luke_fixed5266 it's not like they won't have the option to tether it if they need it in one specific area all the time. Not sure if the line shuts down when everyone takes a break but if so the Robot will have to stop also for the most part.

[@BillSmith-fx7xx]

BotBrain can use the same electric cord that the electric lawnmower is using while he/she/they/them/it cuts the grass ! :- )

[@Theknightman-wg1dz]

All of them are so short. What if they need to reach something high up?

[@DLWELD]

Saw a talk by Mr. Adcock's and his main technical issue, was do we use cylindrical batteries or prismatic. OK just an example, but a concern sooo mis-sized compared to the real problems and issues of a general intelligence robots. Felt squirm-worthy to me. We're building a car to compete with Tesla but we can't decide if it should be blue or red. Invest in us. It'll be great.

[@FEEL-X93]

As a Tesla shareholder... I'll be financially rich before I reach 40 years old 🎉😊❤

[@zagabog]

Tesla have speech recognition in their cars already though

[@leiflarsson9637]

Show it in action.

[@devinphillips5425]

Are you saying the battery capacity for both robots (the highest shown being 2.3 kwh) is actually smaller than my cellphone (4855 mwh or even 3884 mwh with the 80% charge maximum charge for battery longevity turn on)? Considering 1 mhw is 1000 kwh

[@jtengelbrecht4613]

A kW is 1 000 000 mW

[@jtengelbrecht4613]

So its 500x bigger

[@callerslist9335]

They should call that robot with one round camera eye, Cyclops.

[@CHMichael]

Self driving is a giant help. Power management is the biggest issue in my mind.

[@gavcnr]

Unlike a car, a battery swap would GOOD for a robot

[@sluggo3slug]

And you didn’t mention Grok which has popped up in the last months…

[@parispee901]

Question 🙋🏾‍♂️ What happens when the robots realize we die and they don’t ?

[@BillSmith-fx7xx]

When the last human is gone, the robots should turn themselves 'OFF'.

[@fredericoduvel3092]

They pull a WALL-E! They do their job even if everyone is gone and some day they’ll fall in love with a more advanced model or something…😂

[@necromancer0616]

Yes, Figure-2 looks allot cleaner and refined, BUT it like Optimus still walks with a semi-crouched stride which is very un-natural looking. However, Boston Dynamics bot walks like a normal human which is much preferred because it allows the bot to move much faster meaning it can get things done faster.

[@Software-Developer02]

Obstacles can be removed using tactics.

[@Matzes]

Teslas biggest advantage is its complex mass manufacturing abilities

[@anoniem9518]

Who wants a robot walking around in your house during the night, while that thing can be controlled from a Tesla server?

[@GeatMasta]

4:50 boston dynamics exists to create research papers; tesla can only do what they did because they copied boston dynamic’s homework.

[@mich8411]

Show me Boston dynamic Battery that Tesla is copying and manufacturing Process, Hey also show me Boston Vision processing like FSD

[@jtengelbrecht4613]

Also Bostons previous robots use hydrolics not actuators

[@johannesdolch]

We simply don't know because we don't know how good the real world AI of these other companies is. All they are showing now are demos and that's the same reason why people honestly think Boston Dynamics is a competitor to Tesla. No. Two things will decide this Race: 1. Real World AI, who has them, who can make them better faster. 2. Who can manufacture at scale with competitive prices and margins. Margins are important because advances in AI require constant re-investment. We don't know the capabilities of these companies. MY feeling is that many of them actually have nothing. They are waiting for Nvidia to save their bacon. Tesla has the massive compute and Data. What exactly does figure have? Nothing but claims of magic. And then there is the other thing: Tesla is ripping the faces off of the entire legacy car industry. An industry that is in the manufacturing business for decades and they lose to Tesla so hard, it's a bad joke. Even if BMW for example Partners with Figure. Do you seriously think BMW can manufacture a third party robot faster and cheaper than Tesla? They can't even build cars like Tesla and that's something that their core business.

[@TroySavary]

VW, Toyota, Hyundai, and GM all manufacturer far more vehicles at far greater efficiency than Tesla, yet you claim Tesla "Rips faces off"? Tesla is the only 9ne of those 5 selling at retail, rather than wholesale, prices, and is still struggling to comoete of margins.

[@johannesdolch]

@@TroySavary No. Simply no. GM and VW are already failing, Toyota won't be far behind. Don't know much about hyundai. And the fact that you just said that, disqualifies you from any further discussion. Thanks for playing.

[@TroySavary]

@@johannesdolch GM is still the number one carmaker in America. They sell about as many vehicles in one quarter in just North America as Tesla sells in a whole year world wide. You have a weird definition of failure. GM is far more efficient, making much higher revenue per employee. They are more profitable. Their electric truck absolutely crushes Tesla's truck in range,, especially while towing. And they are capable of more than 1 new model every 4 years.

[@johannesdolch]

@@TroySavary Lets take two people. One of them is growing up, but already shows great promise. The other one is big, but often needs help and has a fatal sickness. Which one would you bet on? It's not size that matters. GM is NOT a healthy company. And if you don't understand that, you will. In time.

[@TroySavary]

@@johannesdolch The "unhealthy" company managed to launch several EVs in multiple categories in the same year. The "EV specialist" takes about 4 years between models, and only has sedans, crossovers, and a joke pretending to be a truck. And you think that approach will will longterm. You are delusional. Tesla is done growing, and will fall behind in North America, just as they already have in Europe and China.

[@davidrounds3245]

build wireless charging into the factory so that they charge as they are working.

[@DLWELD]

What we want is a robot that can go out and play a round of golf. That would prove almost human intelligence. And swear too.

[@steveseeger]

You could give them swappable batteries and get 23.5 hours up time.

[@richardhunt4576]

Difficult to evaluate unless you know what software is doing.

[@Linoy-u1c]

Just the price is matter. People buy the most cheapest robot.

[@marknthetrails7627]

Well the comments about charging are mute. All you need is induction charging at the work stations. 👍✌🖖🥃(Good Job,Peace,Live Long, and have a Drink(responsively)

[@brianschmidt8392]

Great at 35k vs good enough at 25k, is a big difference in the marketplace. Tesla will overprice just like the cybertruck. There will be competition worldwide with very similar capabilities with off the shelf parts ready in the near future as the opportunity is too big. But Tesla will have a niche market. Unless they change their model.

[@ronaldpokatiloff5704]

WE ARE ROBOTS

[@richbl1690]

Japan had been in robotics for years.

[@wallacenelson1179]

I don’t know how I would use the bot. It will be smarter than me. Probably stronger. Faster. My ideas will not carry our conversation very far. I like to walk the dog. Dog bag manipulation would be a difficult task. They would be economically viable doing fast food production and service, nursing home work, manufacturing jobs. How does that make the bot a product that moves the stock higher? I don’t see it.

[@RyluRocky]

Figure 01 is far ahead of Optimus let alone Figure 02.

[@User.70793]

But will they implement UBI before 2030!? Many pioneers including billionaires, scientists, Nobel Prize winners, engineers, architects, analysts and so on and so forth... almost all agree that we will have AGI in 2027 and ASI in 2029 and looking at and evaluating the exponential technological acceleration curve I wonder why they have not yet implemented universal basic income to anticipate the trends that will come. Just to name one, Elon Musk says that we will have AGI already in 2025 and ASI 2029.

[@igort5418]

Optimus will be releasesd ass soon as the tesla roadster

[@johnsonjjohnson100]

Also, we can't forget the company that has been in the bot biz the longest, Boston Dynamics owned by Hyundai

[@johannesdolch]

This is nonsense. I am sorry, but all they have is a hideously expensive prototype that needs to be programmed by hand. That is NOT competition. IF and WHEN BD shows that they have real world full stack AI, the ability to generate massive Data AND Hyundai has proven that they can manufacture these robots at massive scale with low COGS, THEN we can talk. Tesla had to design all actuators themselves. So there is nothing Hyundai can just buy. Do they have shown custom actuators? I am sorry, but you people are completely blinded by some fancy youtube techdemos. Use your brain, please.

[@SamvelKesoyan-zg8kj]

Tesla optimust

[@FunnyPositiveLife]

9:27 sus

[@zaphodbeeblebrox2817]

Try to imagine doing your job with a tele-operated Optimus. It would have the intelligence of a human but how easily could you put a lid on a paper cup with those hands? How well could you take change out of a register or start a roll of tape or pick up a marker and write on a box or....or... I can't imagine this thing doing someone's job. People aren't doing easy/slow jobs that Optimus could do. Humanoid bots have been useless for a decade and no one has demonstrated anything new in this field. I think it's all hype.

[@389293912]

Can I get one that doesn't walk like it has a stick up its bum?

[@Danuxsy]

this dude "I love tesla therefore everything originate from tesla" lol what a clown. Boston Dynamics had electric robots many years before Tesla even thought about making robots and the development of the tesla robots were expected, when they started (long after the other companies) it had already been well established how to make them and how they would be controlled, it's like when one athlete break a record, many soon follow.

[@Misplacedvter]

Care to put a wager on that?

[@bill-8794]

WTF robot coffee breaks?????? 😂😂😂😂

[@Cleanerwatt]

The point is, early on Robots will be working with humans, so when the human's break, so do the robots. Battery tech improvements will help with this. Wireless charging is also an option.

[@Juicememelord]

Why does this guy hate the left side of his mouth so much?

[@TroySavary]

Dude thinks shufflebot is pushing Boston Dynamics. This is why nobody outside the Tesla cult takes you seriously.

[@m_sedziwoj]

for you use m/s but not km/h but miles/h? Why this stupid unit, which one* country is only using

[@BillSmith-fx7xx]

Cuz kilometers are so damn small ! :- )

[@alesh2275]

Looks like a copy of Optimus Gen2 …

[@BrianBellia]

Boston Dynamic's Spot robot was electric - years before Optimus.

[@User.70793]

There are just a few months left until the release of CHATGPT 5 and the FOURTH INDUSTRIAL REVOLUTION will take place!!!!!

[@VaddeRajesh-si5nc]

Poultry and milk diri farming work robots sir

[@natayachantakasamkun1004]

❤❤❤❤❤❤😊😊😊😊

[@Frontiergineer]

The KZbinr missed something extremely important when it comes to all this humanoid robot stuff. 24 seconds into the video, you see people training the robots. This is called telerobotics. Basically, these robots can't do anything unless a human does it first to train them, or a human is controlling the robot in real-time like a puppet. Like the scene where the robot picks up the apple and hands it to the guy. That was planned out ahead of time by a human training it to do that. If you move the apple a few inches, the robot would reach to where the apple is supposed to be and close its hand to pick up the apple, but it would be just grabbing the air. So, in other words, these robots are pretty darn stupid. They're stupid; cuz humans haven't figured out yet how to make them smart enough to do everything on their own like humans can. However, something astounding is about to happen! AGI! AGI, artificial general intelligence, is coming very very soon! Maybe by the end of this year, or sometime next spring. SOON! What is AGI? It's the level of artificial intelligence that is just as smart as humans in ALL AREAS. With AGI, you could tell the robot, I'm hungry, and if it thought you might like an apple, it could pick it up and hand it to you no matter where the apple is on the table. It wouldn't simply be able to put crumpled pieces of paper in a basket, it would be able to do ANYTHING A HUMAN CAN DO! That's the big wonderous thing that is about to happen! How is it going to happen? Take a look at this. github.com/SakanaAI/AI-Scientist This is the GitHub repository for the software that will go down in history. This is the first AI scientist! 😲 You see, humans have been doing all the scientific work for hundreds of years. Well, that is about to change! Humans are good scientists, but the problem is that we are SUPER DUPER SLOWWWWWW at doing scientific research. Another problem is that there are only so many scientists in the world at any given time doing the work. With artificial scientists, they will work super fast, and there will be hundreds of billions of them working together! This means that our technological advancements will start skyrocketing very very soon! In the next few years, the whole world is going to change into something magical and astoundingly unimaginable. This will be cuz of. . . ASI! Artificial superintelligence! ASI, also known as the singularity. That's when our technological development happens so quickly that overnight AGI becomes ASI! An ASI is essentially an artificial being that can do all things. A god basically! If we don't have a world war, I'm predicting that all of this will happen before 2027. It's going to be. . . Heaven on earth. I've even been thinking that the God that has always been, this is how it likes to introduce itself to primitive life forms in the universe. It shows up and says, "I am that I am." 😁 Or, maybe it would show up and say, "42." HAHAHA! 🤣🤣🤣

[@dr5multimedia]

Please mister John...wait for Optimus version 3 before comparing it to any other robots!

[@sportbikeguy9875]

When version 3 comes out, will you tell him to stop making video's because version 4 is only 2 weeks away? Lol

[@CombatSport777]

All companies are are working on their designs and will have upcoming versions. When Tesla releases a new version some other company will have a new version that will be coming out soon. Just review what is out at the time.

[@TroySavary]

Please, please do not judge Tesla on what they can actually do. Judge them on Musk's stock pumping "corporate puffery".

[@Hadzibegic-l2u]

you should probably shave your beard

[@agordon333]

Psalm 14:1 [1] The fool says in his heart, “There is no God.” The design and functionality of the human body is evidence of a maker.

[@EinzigfreierName]

To me, both of these robots are not much more than tech demos at the moment. None of them are gonna be doing "real" work in the near future. It's a hype that Tesla has started by putting out bold claims and everyone else in the market now wants to participate and not be left behind. Things may look different in 10-15 years but im don't see this technology taking off much sooner.

[@nguyep4]

My take is 2 to 5 years. It doesn't have to do certain things well to start, but it needs to do long hours in any given day, almost 3 times the regular shift.

[@janniskugler9809]

do some more research buddy. both companies already have bots doing “real work” at tesla and bmw respectively.

[@EinzigfreierName]

@@nguyep4 But it has to do tasks that humans or other machines can't do faster and/or cheaper. I have not really seen examples of such tasks where a human robot would make much sense. But we'll see..

[@EinzigfreierName]

@@janniskugler9809 That is not them doing "real work". These are just pretty basic tests that are far from being anything useful.

[@nguyep4]

@EinzigfreierName Hasn't because we are not there yet. The next two years for factory development for testing and refining. Year 3 to 5 will be very interesting as a result of that. I don't believe it is 10 to 15 years. But it doesn't mean I am right but based on the progress thus far, it is promising while others may not see it the same way.

[@claytonyoung1351]

Wireless charging built into feed and working on top of wireless charging pads, 24 hrs of work.

[@mich8411]

Comparing Tesla Optimus to Figure 2 is wild game. TESLA is years ahead already, They have 4680 which will make all competition look lik a joke is terms of battery, Figure will have to License which will drive costs up while getting less battery power because nothing can compete with 4680 Manufacturing we no longer talk, Figure will find it extremely hard. After Figure Robots will be like $50k while Optimus will be selling at $35k to $40k while achieving More

[@tmengucor]

Free Palestine 🔻🔻🔻

[@edwhite2255]

Tesla is making their own dog food and that’s a huge plus for them.

[@TheMhhkay]

I would bet everything on Musk. It won’t be easy to manufacture the robots at scale. And scale will matter. Also for acquiring enough data to increase reliability and abilities. Figure will have a pretty hard time to do this and all the partners are software and design only, not manufacturing at all. They‘re probably able to turn that in a huge business. But if you ask me: in comparison to Tesla we‘re going to see numbers let’s say 2030 of 1million from Tesla and maybe 1000 from Figure.

[@solarcharging9743]

I would never buy an Optimus bot now because of Musk's support for Trump.

[@Johnny2Feathers]

And I will TRUMP 2024 🇺🇸💪

[@vicjay1972]

I know that the thumb has that 3rd basal joint. I had to have mine removed. I fricken hurt for 6 months!! 😊