No humanoid robot should cost more than or equal to an cheap car. Even though it is A.I., the actual cost per pound in materials is much less in the robot.
Bot on mars first is a must, 1) set up solar farms 2)set up housing 3)set up equipment to drill for water. 4) space mining. Bot skill training is where the money will be!
Imagine fleets of tireless robots that have a per labor hour of $1... hired to do such jobs as cleaning up rivers or toxic environments, or dry kindling from forest management tp prevent annual forest fires. .
Construction companies would be interested in these robots for laborers. lowering the cost of construction. In the beginning it could be lease, then lease to own.
It's nonsense. Tesla's investments in AI is so far behind the competition that if you actually looked at the numbers you'd wonder why they say Tesla is even in the race. And I've worked with robots for nearly 30 years. Robots aren't taking anybodies jobs and not saving and companies money.
Totally cause these robots won't break down need maintenance and cause delays to work. We are not going to see robots in construction maybe in the far future but not anytime soon.
The more robotics and AI progress the more jobs will be taken away. Soon(within 50 years) an universal basic income like system should get implemented or most people wont be able to work to live.
I will teach my Optimus Prime to rob the Jewelry store down the street. I will dress him as a gang member for disguise. I will also equip him with deadly force to keep me safe from the badies. I will rule with my Robot.
Agreed, but they only need the least best possible way because the work day goes from 8 hours a day to 24 hours a day. It don't have to be perfect, just good enough. Be safe out there.
Hardware will change more than most of us can imagine. AI will allow autonomous operation of machines at scales and in time frames most of us find hard to concieve and potentially using senses we do not possess. Today, interfaces are mostly built for humans. As AI evolves most of todzy's interfaces will be replaced by interfaces built for AI. Low level control architecture will be optimized for both local and remote AI control. Much of the machine world will be transformed in ways few have imagined. Consider the dial phone partyline user from the 1960s reaction to the new iPhone 16 (hardware) and its capabilities and resources (hardware, software, and AI). AI will likely have a greater impact on technology (hardware) and on society in your life time. What would be your reaction if you could see those changes today?
@@tomcurnett7519I want one to provide care support for my wife and I, as we are both disabled and have to rely one family and friends to do all sorts of chores that we are unable to achieve. Hope there’s finance options whether lease to buy, or something.
Bottom line... We are talking about a machine that can be built for $10k. $300 a month for a service agreement, and 3 years of use - $20,800. 350 days at 20 hours per day for 3 years. 21,000 hours. Basic labor will cost $1/hour.
You’re right that labor costs will plummet. But it will plummet to the point robots make massive amounts of money. 7000 hours a year times $20 is $140,000. After 1 month, Optimus is paid for. If you bought 10 Optimus bots, you’d make $120,000 a month or almost $1.5M a year. Optimus will change everything. We will be fighting to buy them.
In a nutshell, yes. For a couple of decades it will need to work toether with highly capable humans so it is no like it will eliminate ALL jobs. It will eliminate some, create others and things will change. For the better or worse is hard to say at this time.
IMO, non-humanoid robots will ultimately be far more capable and useful than robots in a human form factor. A non-humanoid robot possessing multiple limbs, sensors, etc (i.e. more than a human) that can transform itself to optimize itself to suit the operating environment and tasks it must perform will be of far greater utility/benefit than a humanoid robot.
Problem with this is if you replace all the workers with bots then no one will have an income stream to buy the stuff the bots make... kind of self defeating. space exploration and other dangerous activities makes sense.
@@SparkySho not in human nature to make things for free.... What business owner is going to invest millions in bots so they can make no money and give things away virtually free?
@@CathyMartin-en2tl tell that to all the car production line workers that lost their jobs to robots. That scenario was 100% about increasing profitability and not paying a work force or having to allow workers rights etc.. Now see if the owners will still do that when there is no increase in profit and tell them they have to contribute to a universal wage for everyone to sit and do nothing....
Very improbable because it would imply that everyone, grownup or child, could afford monthly payments for the robot in addition to the cost of having shelter and food.
As far as demand for robots here on earth. If you could buy an Optimus robot or two or three that could cook and clean and take care of you so that you don't have to move into assisted living, which is REALLY expensive for one you'd enjoy, it would be significantly cheaper (less expensive) to buy an Optimus robot or two or three as assisted living that you wouldn't hate is $5,000 a month which most people can't afford. Even at $20,000 for an optimum robot would still be a lot cheaper than a "nursing home" where basically the person would be laying in a bed and would need general and mostly easy medical care plus house cleaning and feeding of the person. Also keep in mind that Optimus robots have 4k cameras in them (or can) and that would allow doctors and nurses to remotely give a good and easy overall examination of a person's whole body REMOTELY. This would make it SIGNIFICANTLY less expensive then hiring someone to get the person to the doctor if it is just for a checkup. And if the person is semi mobile then the Optimus robot would be programmed to be caring and wouldn't have good and bad days. And it would be strong enough to help the person in any way that they wanted. Plus they wouldn't have certain work hours or families to go to nor would they need sleep. What about charging? Simple! The Optimus robot could literally plug into a cable which would be long enough to take care of the person in their bedroom and could assist the human in anyway needed as long as it was at bedside. And by the time the person needs/wants to get up, the Optimus robot would be fully charged. it would only take 15 minutes to have a pretty good charge and way less than an hour to get a full charge. But they could plug it at any time of the day or not they aren't needed with a cable that would be long enough for a area of maybe four to five meetings which is about twelve to fifteen feet. This care would be SIGNIFICANTLY less expensive than any assistant living or even nursing homes until the later part of the person's life where they are very, very near the end of their life. Then an Optimus robot could use a Tesla car to get them to a doctor's visit or to a hospital. Actually I expect that Tesla ambulances will be available at some point specifically for the use of Optimus robots that need to transport a person to a doctor or to a hospital. I think they should start thinking about sooner than that later as within eight years I could see that being another segment that would pay off for Tesla. They would need some kind of electric van for that which wouldn't be hard for them to design and build.
Exactly, Elon thinks these will be selling for 10k when they get manufacturing up to scale. Much less expensive then a nursing home. If it can do everything a human can for a one time cost of 10k. That is the end of nusing homes.
Some good thoughts there, but as far as calling in for a remote doc visit, you are living in the past. AI has already surpassed human docs in diagnostic abilities, and it will be so inter-meshed with you daily activities that you will never have to see a doc unless you need a procedure, and even that will probably be done by a robot during a house call. Your health and condition will be continually monitored by brilliant systems at all levels, thousands of times an hour. You will be immediately notified of any changes that need attention.
The key to robots will be software / AI. The physical side is done for basic human tasks. Tesla has AI and mass production expertise - I don't see the same in any other company.
Tesla will probably sell app like software packages for different functions, and focusing on this area would probably be worth more than the bots themselves.
Industry will embrace the robots where possible and eliminate jobs when given the opportunity in order to maximize profits. Without jobs, people won't be able to afford to buy robots and universal income will come into play. There needs to be a fine balance, otherwise we will see a even greater divide between rich and poor. US companies gave up manufacturing to China and surely we will give up jobs to humanoid robots but at least we can bring back manufacturing to the US.
The subscription model has a glass ceiling of the minimum wage. And I don't think it will make any sense to replace a human unless the hourly cost for a robot is at least 50% cheaper than hiring a teenager or outsourcing it to Vietnam. So charging a premium on these robots will be forever out of the equation, and so it makes no economic sense to think that these robots will replace cheap labor in a global market lol Imagine if Netflix pitched this magnificent business proposal about never raising prices ever, stuck at charging 7 dollars forever ever, who on earth would invest in that? What that financial adviser is saying about eventually raising prices as their capabilities are improved, it is just ridiculous. As a low level cheap labor I need it to do one programmed task well, move this box from here to there, check inventory, scan items. That's it. I don't need more enhancements, and I will not pay for any other upgrades ever, I don't need it to jump or even better cognition. The best paying jobs are managerial, decision making and creative positions, and those don't require physical robots, AI will be humming quietly in a server room for those roles. You don't need a body for that. Plus, imagine having a whole plant paralyzed for a bad update on all your "workforce", freaking nightmare. The whole Optimus thing is doomed to fail: for industrial uses, you need specialized machines not human like inefficient and ineffective anatomical copycats rofl. People will always hire a cheap immigrant to clean the house, not a 30K USD bipedal roomba with less dexterity than a 3 yr old, what are you guys smoking
Why would there not be “apps” for different functionality that can be subscribed to? For example, Joe wants his bot to be a gourmet chef, an artistic gardener, and a math teacher for his children, so he leases his bot from Tesla and then goes to the Tesla App Store to subscribe to these apps. Independent developers could create apps for Tesla bots just like others do for iPhone apps in Apple’s App Store. An electrical contractor leases 20 bots for use in his company and gets 20 Master Residential Electrician App subscriptions. A physician leases bots as pediatric “nurses” and ‘physicians” for his clinic (or more likely, each family subscribes to a “pediatrician” app for their home robot). In this kind of situation most of the money would be in app subscriptions more than the bot itself.
Yes. Tesla has an incredible headstart with their robot car fleet that can translate quite easily I would imagine to other incorporations such as humanoid robots
I appreciate the analysis, but it reminds me of how politicians justify placing a tax on certain activities; that the affect of a tax will not reduce the activity as people and businesses avoid the activity in order to avoid the tax. In the end, the envisioned tax revenue isn't realized because it made the activity too expensive. Similarly as bots are added to the $16/hr labor market, human laborers will be freed up, causing more competition among laborers.. Some will work for less than $16/hr in order to find work and compete against the robots. Some human $16/hr laborers will upskill to find the $18 and $20/hr, driving those rates down because of the added competition. Ultimately, I think the cost of labor will significantly be reduced, diving to near $0. These costs outline in this video will slide as the affect of robot automation creates lower cost workers.
The only tax that's needed is a tax on the income made by the robot producers. Why would we tax the bot itself? We currently have millions of robots in factories already, and they aren't taxed. They are considered equipment. Just because a bot can walk shouldn't change the tax structure.
His subscription and sale prices are a complete fail because he is not accounting for competition. If he charges $300k and the Chinese charge $15k I will buy zero from him and three from the Chinese.
The concept of a humanoid robot for industry it is actually stupid. You don't need "generalist" robots, you need highly efficient and specialized robots. Anthropomorphic robots for business makes no freaking sense, absolute zero sense. Boston dynamics have very interesting designs with legged-wheeled robots which I can totally see them being useful in a warehouse. Bipedal and humanoid? That's really stupid unless they are used for interaction with humans, like companionship and monitoring of elderly care.
@@xantiom i think you still dont understand why humanoids make sense. You see, when it comes to robots, hardware problems have pretty much already been resolved for the most part, the main issue holding back robots is the software. Boston robots may be able to do a lot of programmed hardcoded fancy moves, but it takes a lot of effort to code it to resolve new kinds of problems. For optimus, the main goal is to resolve the software hurdle. how? By imitating humans via data training, lots and lots of data training, because humans can resolve a lot of very complex problems, you just need to copy and paste and let the robot imitate. Saves way more time than coding done by boston. So if you build a humanoid robot, data is freely available to copy and paste without much need for altering the code. Non humanoids dont have a ready available source of complex data to copy from.
I think Cern Basher is wrong about Tesla Robots being off planet "later" and here's why. I think that sending humans to mars in quantity is going to be SUPER risky. And keep in mind that if we WANT to do that, we need some place for them to live. There is a company that has used a regolith substitute to 3D print buildings but you have to get that onto the Moon, you have to assemble it, and you have to take real regolith and supply the 3D printer with that material. I think it would be SIGNIFICANTLY safer to send Optimus robots to the Moon and operate ALL of the equipment to do EVERYTHING. Literally everything. It would be easy to upload all the different "jobs" that Optimus robots need to do to Starship either before Starship leaves earth or earth orbit or even when in earth orbit and then have Optimus robots "plug-in" like Borgs did in Star Trek (though with less cool looking hardware) to get their programming. Keep in mind it will be SIGNIFICANTLY less dangerous to get them into space, get them to the Moon, get them ONTO the moon and then have them assemble the 3D printer that will print the buildings for humans. They could also assemble everything they need for themselves from charging stations, to vacuums that would vacuum Moon dust out of their parts, etc., etc., etc.. And these Optimus robots could drive whatever equipment you have. They could even be programs to "operate" on each other as parts get destroyed do to Moon dust which is very corrosive. Also, most of what exploring can be done SIGNIFICANTLY safer with Optimus and they would be much easier to house since they don't need oxygen or anything else to "breath" and the only thing they need to feed on is electricity. So easier, easier, easier. And that's just the Moon. I believe the project on Mars where Elon wants over a million people there, which is what you would need to be able to build a city and have enough humans there to protect us from ourselves or some unknown "thing" like a lot worse COVID or a lot worse asteroid than what killed little Dino from the Flintstones.
exactly - Tesla will need to reduce e.g. a Martian city to the absolutely minimum bootstrappable seed that can be loaded onto a fleet of starships and sent to e.g. Mars where it can start with the acquisition and refining of raw materials, then going up a ladder of producing more and more elaborate tools, machines and factories until you have the components you need to build lots more Optimi who can then build a city and do some terraforming before the Humans actually arrive. Except that Humans will probably be involved in the process the first few times until it can finally by shrinkwrapped and scaled at which point we can literally seed the Universe as each seed's job will simply be to produce more seeds (a seed includes a Starship) - just like a tree.
I started 8 companies over my lifetime This is a game changer I would have absolutely saw the opportunity to make this my new business opportunity Absolutely a no brainer
The debate has never been over how valuable an Optimus robot would be if it were as capable as Musk implies it will be. The debate is over how long it will take to get there - months, years, or decades. Full self driving has taken far longer than Musk predicted, and it's still not at level 5 autonomy. The AI problem that needs to be solve for Optimus is far harder than that for FSD.
Who will afford a robot when they take all the jobs? It is a tricky thing and life may change more fundamentally than people think. We may no longer have a use for money. Robots make robots, robots mine the materials deliver the robots, no one needs paying.
Sounds like either sell or lease for a base price with minimal skills, and allow 'downloadable upgrades' like being done with the Tesla to sell or rent '$20/hr equivalent' or '$40/hr equivalent' skill, all on top of the 'base capabilities' that can do quite a bit. Even for home use, a base 'cleaning bot' that could also do laundry, and a possible upgrade to a 'cooking' robot, to a 'Mechlin Star Chef' robot skill level might be reasonable, and do the same with manufacturing or construction skills. Each skill could have a time/lease cost option, so you only need 2 hours of a 'cook' daily, but you might need a weekend Michelin Chef' rate for a big party. And different for feeding 2 or 4 or 400 people in that time.
It’s important to bear in mind that Elon is primarily a physicist and thinks in magnitudes. When he predicts 10 billion robots what he really means is that he’d be surprised if it was a magnitude less or more, so he’s effectively giving a probable range of 5-50 billion.
@@thirdplace3973 Yup. Yes - this is what happens if you generate 15 billion more units that can do human labor in 5-10 years. Normally this sort of thing would take hundreds of years.
@@LowkeyXxx it increases also because of fiat currency and printing money out of thin air. Minimum wage used to be $1.25 per hour. 5 solid silver quarters. Today that would be worth $30/hr.
1. Humanoid robots might raise human and industrial powers but only by overshooting the efficient consumption of resources in service of the multiplication of human productivity. 2. The efficient consumption of resources has regulated the operation of industrial firms and workplaces broadly throughout the whole history of industrial capitalism. (Efficiency gains come in concrete steps however so ineffiencies can persist even as newer and more efficient production techniques supersede older more problematic techniques.) Capitalisation serves the goal of lifting productive powers and reducing costs of outputs when those outputs are produced in volume. Surpluses grow because despite increased capital spending the cost of production per unit of output still gets reduced/economised. Similarly, as surpluses/profits increase these can be recirculated into further capital expansions so long as this can be done efficiently. 3. Only efficient production (usually implemented in a phase of capital expansion and with a pay-off over the long term) gives one firm a practical edge over another. If humanoid robots don't lift production - they don't - they are utterly useless. Currently, the advocates of humanoid robots are promoting the bug that they don't raise productive powers (at a competitive capex investment) as if it were a feature. Companies that go down that path will quickly find themselves in financial trouble. We were already on the right path with robotised machinery and tools that increase the speed of production and reduce per unit costs. This humanoid robot thing is a complete non-starter for industry. 4. Non-humanoid robots multiply human powers efficiently because they increase the speed and efficiency (reduction of resource requirements) of production. Humanoid robots don't do that and any business that spends its profits on expanding uncompetitive/inefficient/overpriced productive output will fail. 5. Humanoid robots only make financial sense as a replacement of human workers without any compensating UBI. And, that is a crap idea.
The uncertainty I have is about how intelligent Optimus will be. That's how easily it can learn new tasks. Simple repetitive tasks will be enough at the beginning. But more flexible ones, like household work, is possibly out of reach for a long time. This can easily cut the profitability by a factor of 10.
If a Tesla with FSD can drive itself then I think an Optimus, with software many generations on from the car, will be useful. Many people all want the same domestic tasks, so once an Optimus can safely navigate around domestic settings it won't be long before each s/w release (monthly, weekly ?) brings you new skills, just like is currently occurring with the cars (c.f. recent release of Summon capability) : - vacuuming and cleaning - washing and ironing - cooking - mowing and gardening - general DIY - more specific DIY - carpentry, plumbing, electrics, heating... - ... And this is assuming s/w releases being structured like Tesla cars i.e. in serial but in order to scale, development will have to become modular and proceed in parallel like an AppStore with multiple companies producing training data in their particular area of expertise ... In fact you will probably have a whole selection of e.g. Cooking apps just like you have a selection of different restaurant types - Indian, Chinese, Thai, French etc.. It is going to be an amazing time to be alive !
You can't assume that Tesla is going to corner the market for those robots. It seems to me that Unitree and Figure 02 are already superior to Optimus. It remains to be seen how they will compare in usefulness or price. Optimus will be one of many available options.
The central question remains: what happens to displaced humans. Watch Elon and Sam Altman discuss this a few months ago and agreeing that first to go would be blue collar jobs, then white collar, then creatives. Software developers are already getting squeezed by AI. This guy shows no concern or appreciation of the depth of the problem. Talks about replacing all labor. Not a single word of concern, mindless drivel about trillions of dollars. A traditional answer is that new jobs will emerge - but these would be also taken by robots. Humans would have no jobs if robots are so capable.
I think the answer is that post-robotic humans won't have "jobs" and certainly won't "need" them as essentially for survival. What humans have and still might have are "responsibilities, duties, goals, desires, etc." which will fully occupy their times. In other words, humans "will get a life" in trade of "jobs" and "working for the man".
@@darylfoster7944 Not really. There are lots of what used to be manual labour jobs that are now accomplished by machines. Milking cows and telephone operators are an example. There are countless more.
I think the lease vs. buy argument is wrong, I suspect it will be more along the lines of high end CAD systems, you have to buy the robot outright, then you have to pay a monthly maintenance fee. They might require the maintenance fee also, not make it optional SO you get both the sale and the monthly fee.
However you have to understand the gross number on human robots is the total market, but Tesla aren’t the only ones building human bots, so Tesla’s potential value will depend on their market share.
@9:39 You don't understand the products. There's the robot(s) and then there are the skill(s.) One robot can master many skills. First you lease (or sell,) the robots, and then you lease (or sell,) the skill(s), and then you lease (or sell,) the equipment to train the robots in new skills and then you lease (or sell) the ability to do OTA updates to specific robot units.
The calculation doesn't work that way because there is cost pressure from competing products. In the end, such a bot must and will be much cheaper and this will promote its adoption. Competing products will probably be more successful because they are already more developed than Optimus.
I actually don't know if they will ever go to a sales model. The brains will constantly improve, and ultimately, they are selling capability. Probably easier and more profitable to lease them out... and safer. I can think of other reasons to stick with a rental model: 1) Waste/Recycling control: It would be bad if old robots weren't disposed of properly, from optics to environmental to regulatory. 2) Unauthorized Repurposing: Could you imagine if these things were turned into some warlord's personal army, or into gangland enforcers. 3) Competing with old models: Why give customers the choice to save by sticking with old stuff? 4) Data Collection: You want as much as you can get. It's the life blood of an A.I. company. 5) Control: Why give any up? These are just a few things I can think of that say a sales model is bad. Maybe competition will force your hand and you'll have to sell, I can't say for sure. I just know I'd stick with a subscription model for as long as it was viable.
Most of the humanoid robot labour market doesn't yet exist. Once they are made en masse and work well, there will be more jobs for them that humans DON'T yet do than currently existing labour.
5:09 Imagine how much effort it would take just to build 1 billion robots alone. Tesla stock would have split 5-6 times by then i’m sure. Maybe 10 times once they ramp up
Have no idea when these robots will be done. 10 years, 5 years, 3 years Can the robots run into fires and carry people out? Can they move as quick as people do
Unitree has their prototype going 6 mps. I think if you want to use them for fire rescue they must change the form factor from human to crab - walk up the side of the building and carry the fat lady back down.
Yes world wide there may be 14 billion phones in use, but the majority of those are low cost phones. Unless the price for owning or renting/leasing a robot comes down to a few thousand dollars selling price or 100-200 dollars per month for rent/lease, I don't see a majority of the population having a personal or even family robot. I can tell you now, that my income would not be enough to even be able to afford those prices. And having robots take over a lot of production jobs, would also mean that for a large portion of the workforce, their jobs would be on the line, meaning no income and no way to afford a robot. Unless of course a form of a decent universal income would be set up and I don't see that happening in the current capital doctrine that runs the western world. Sadly I feel that the way society is run at the moment, robots will take over more and more of the lower skilled labour. And with every increase in capability, the next tier of jobs will cease to exist. As a result those people will lose their ability to buy stuff and keep the market going, unless that universal income is sorted. If not, be ready for some serious, and I mean French revolution serious type of unrest because people at the lower end of the economy will have no choice but to fight for their survival. In short, I am not sure, those that try and increase the use of AI and robotics have thought through how disrupting this will become for society.
ya i burst out laughing... oh there's 14 billion phones ... therefore theres gunna be 14 billion or more bots! Real high end quality logical analysis hahaha.
Many comments say can't quite replace human labor..yet, Though the robots to Mars to build shelters and greenhouses and Boring company tunnels, is outrageously expensive and probably happening soon
A functional C3PO would be nice to have. Robust automatic speech recognition remains elusive (1-8 meters lips-to-mic, spontaneous speech, multi-talker, etc.) Entropy is also a major hurdle for current AI approaches. A 800 watt hardware-5 cpu might work in a Tesla car but impractical for a functional 3CPO. First princples are being ignored. Vaporware for now.
Food is a distribution problem not a production problem. We harvest more than enough food to feed everybody 3 meals a day. The problem is that the distribution is uneven. More food is delivered to you than the amount delivered to the starving children in Africa.
It seems people mostly seem to think of robots (humanoid, drone, wheeled, whatever form factor) as smart machines. Maybe, your Tesla is and may continue to be just that, even when you can talk to it, like Siri or Alexa. But actually that changes if AI approaches sentience and consciousness. The situation also changes when AI robots are brought into the home or the small business workplace. At that point, they will cease to be like dolls and teddy bears, washing machines and vacuum cleaners, or computers and phones full of confidential data. They will become part of families or business teams, purchased initially maybe but not leased. They can be enhanced or up-lifted, but not switched off or erased or wiped when updated. Domestic applications will need stable personalities. In the meanwhile, smart machines can do dull, dangerous, repetitive work at levels of reliability, quality, and speed far exceeding what humans can do and historically, up to now, have had to do for themselves.
You're crazy to think if anyone's going to pay that exorbitant amount of money when you can buy a Chinese bought eventually for $5,000 so the market will force the sale of the robots cheap
Just imagine the Humanoid robots doing separate tasks and the recurring revenue it can generate? #Massage therapist , #Chef #Cleaning lady #Gardener #caregiver All these could be separate apps downloaded for each tasks.
It's not about hardware - it's about software - no-one currently competes with Tesla on software - Tesla is the only company with millions of [almost] autonomous robot [car]s working out in the wild... - oh and they're pretty good at hardware too.
The main problem with your analysis is it lacks the appropriate inclusion of competition in robot production, programming/training, etc. For example, the Ag industry is now almost fully automated, increasing the human productivity by more 100-fold. The same will be true for all the other areas where human labor's productivity can be effectively multiplied more than 100-fold utilizing robots. Competition will be the primary constraint on the ability to profitably deploy any particular format or specific type of robot, as it should be. Regardless, there will be a huge first-mover advantage in the rapidly emerging marketplace for these super-flexible, human labor multiplying robots.
One gallon of gasoline is approximately equivalent to about 25-30 hours of human labor. Human work 35-40 hours a week. A gallon cost 3.50-4 $ per gallon in USA. So a human physical labor production in gasoline price equivalent to a maximum of 8$/week ( 2 gallons) but cost 800 $/week at 20 $ / hours. 100 to one. Robot cost 20 000 $/ 800$ per week = 25 weeks to pay back the robot. The robot can work 2.5 times longer then a human ( 20 hours/week). So the robot pay itself in 10 weeks minus operating cost.
I think this 20 hours a day thing is wrong - there is no way I would let an expensive piece of hardware sit uselessly for 1/6th of its time. Recharging will be done as it does its job, either wirelessly by standing on a charging mat if the job is stationary or on a cable if moving around a bit. There will be very few jobs where this cannot be managed.
Musk is absolutely brilliant in certain ways. One example is his epic ability to keep the Tesla stock price elevated via a series of "hot new technologies" with "immense market potential" which keeps his devoted followers buying. The smarter traders have left long ago, recognizing what lies ahead. The bot concept is interesting but a close look at what it will actually take to get to markets is sobering. Regulatory approvals for widespread applications is likely a decade away. Safety concerns in human-bot environment are infinitely and regulators know it. But these realities are not Teslas most immediate threat. It's musk's imminent mental breakdown that will bring the stock price down to earth.
If you assume an average of 2 Optimus robots per human and Tesla charges just $1/hour, that's 16 billion * $1/h * 24h * 365d = $140.16 trillion per year of revenue. And that assumes they charge nothing for the robot. Even if you made MUCH more conservative assumptions, no matter how you look at it, you're talking about some truly wacky numbers.
I've been to many poor third world countries. Even in these poor countries they still all have mobile smart phones. These people will NOT all have $25000 personal humanoid robots.
These valuations are egregiously over inflated. If the bot takes $15k to make, we should expect competition to bring the cost down to cost + and acceptable margin, e.g Apples extremely healthy 30%. Thus not a $800k bot, but rather a $20k bot. The cost to own one may grow once governments start to look for income tax revenue replacement. Since industry can use bots to make bots, the smart company should be able to scale exponentially, so I don’t see a supply shortage lasting very long - as long as the parts themselves are not constrained.
"Since industry can use bots to make bots, the smart company should be able to scale exponentially, " Only in theory. You need machines, raw materials, management, capital and so on - nothing has ever scaled exponentially for more than a short period (outside the spacetime expansion of the universe).
He didn’t take into account productivity. These robots are slow because of hardware limitations and power constraints, regardless of how good the AI is. A robot working 20 hour per day will only be able to create the same number of widgets as a human can in 10 hours. The value will be more like $1,300 per month per leased robot.
When they anounced Optimus they were kind of the first and special. Now we already have 15 very hard competitors that seem even to have better bots already now. There will be so many bot companies, the prices and margins will ve very low. Think about it.
I dunno, the big assumption is that a humanoid will eventually be able to do everything an average adult can and I just don’t see that happening. AI-assisted humans? Certainly. Dangerous, boring, repetitive jobs, of course. But I still want a real person (who has a soul like me) to cut my hair, to teach my kids, to make my food, to take care of my elderly parents, to conduct my physical therapy, to clean my house, to coach me through my career development, to pull a splinter out of my finger, to motivate me in the gym, to hang out with. Crap jobs will be automated but there are many more things that only other humans can do for the best results.
As a geek and Tesla shareholder I am excited about FSD's future. Is Elon mistaken when he says that Tesla has an equivalent moat when it comes to applying FSD to Optimus? Tesla doesn't have a fleet of robots being remotely controlled each day like their cars... My concerns are specific to the numerical disadvantage of employees vs a huge fleet, as well as the fact that whilst the driving world is infinitely variable, the goals are highly defined. A general purpose robot faces goals which are proposed to be infinite as well. By my reckoning, this leads to a situation where Tesla starts much closer, albeit ahead of the competition, in a race to an outcome which is much further away, and probably much more limited that the proponents' expectations. As an outsider to the field, I am far from certain of my take and would value any input both from intellectual curiosity and for my due diligence re my Tesla shares.
Also a shareholder. FSD has a massive moat - 7yrs+ head start. Optimus has no moat. It will have to rely on Tesla's access to better engineers, more capital, and Elon's instincts. I would stay invested for both. Upside oportunity is high and downside risk low.
Factor in all the more advanced robots which are already coming to market like the Electric Atlas from Boston Dynamics (Hyundai). This video seems to gloss over Tesla's bot competition.
Home Healthcare, hospitals, Personal robot to cook and clean, do laundry...perfect for older people living alone that don't want nursing homes
Just remember, old age could happen to you.
@@michaelnurse9089
Old age is a great thing considering the alternative.
No humanoid robot should cost more than or equal to an cheap car. Even though it is A.I., the actual cost per pound in materials is much less in the robot.
@@michaelnurse9089 I would prefer a machine bathing me over a person at that age.
@@ChosingGod IF IT ROI SO HIGH THEN NOT GOING TO SELL CHEAP THOUGH OR WOULD JUST BE BACKLOGGED FOR YEARS
Bot on mars first is a must, 1) set up solar farms 2)set up housing 3)set up equipment to drill for water. 4) space mining. Bot skill training is where the money will be!
At this point let's go full robots and zero humans!
@@iko3 we are pretty useless in space, air, water , food, radiation, robot just need solar panels
It will look just like Total Recall, the Arnie version.
Infrastructure living uh huh
Mars project is moronic.
Horrible place to live.
All I need is one Optimus that has learned to make copies of itself.
Brilliant but kind of of trueb
Imagine fleets of tireless robots that have a per labor hour of $1... hired to do such jobs as cleaning up rivers or toxic environments, or dry kindling from forest management tp prevent annual forest fires. .
Construction companies would be interested in these robots for laborers. lowering the cost of construction. In the beginning it could be lease, then lease to own.
Every company ****
It's nonsense. Tesla's investments in AI is so far behind the competition that if you actually looked at the numbers you'd wonder why they say Tesla is even in the race. And I've worked with robots for nearly 30 years. Robots aren't taking anybodies jobs and not saving and companies money.
Totally cause these robots won't break down need maintenance and cause delays to work.
We are not going to see robots in construction maybe in the far future but not anytime soon.
@@avenged7foldgamer MARS FIRST!
And they will NEVER lower prices of the end product.
Imagine these things restocking walmart or your favorite grocery store
Imagine these things in the army - no recruitment problems. China will be right on it.
@@DLWELDreal robot warriors won't be humanoids.
That will be their first job.
@@irri4662 Crab is theoretically the best land form factor.
The more robotics and AI progress the more jobs will be taken away. Soon(within 50 years) an universal basic income like system should get implemented or most people wont be able to work to live.
And don't forget robots for war and crime
Damn that's cold. Possibly true.
Solar bot t-shirts .
War is obvious but crime I never thought of before. Robot theft itself may end up being bigger than the slave trade was.
And warcrime
I will teach my Optimus Prime to rob the Jewelry store down the street. I will dress him as a gang member for disguise. I will also equip him with deadly force to keep me safe from the badies. I will rule with my Robot.
Hardware will never be a problem, AI will learn how to control the hardware in the best way possible.
Agreed, but they only need the least best possible way because the work day goes from 8 hours a day to 24 hours a day. It don't have to be perfect, just good enough. Be safe out there.
Hardware will change more than most of us can imagine. AI will allow autonomous operation of machines at scales and in time frames most of us find hard to concieve and potentially using senses we do not possess.
Today, interfaces are mostly built for humans. As AI evolves most of todzy's interfaces will be replaced by interfaces built for AI. Low level control architecture will be optimized for both local and remote AI control. Much of the machine world will be transformed in ways few have imagined.
Consider the dial phone partyline user from the 1960s reaction to the new iPhone 16 (hardware) and its capabilities and resources (hardware, software, and AI). AI will likely have a greater impact on technology (hardware) and on society in your life time. What would be your reaction if you could see those changes today?
Ross Gerber :
"nobody want to buy a Optimus" 😂
Ross Gerber is the borderline idiot with a grin on his face.
For this 75 year old man, I want a bot as an exercise buddy. Teach them to play catch with a medicine ball and I'm there.
@@tomcurnett7519I want one to provide care support for my wife and I, as we are both disabled and have to rely one family and friends to do all sorts of chores that we are unable to achieve. Hope there’s finance options whether lease to buy, or something.
Ross is a 🤡
Wouldn’t you prefer real human contact?
Bottom line...
We are talking about a machine that can be built for $10k. $300 a month for a service agreement, and 3 years of use - $20,800.
350 days at 20 hours per day for 3 years. 21,000 hours.
Basic labor will cost $1/hour.
You’re right that labor costs will plummet. But it will plummet to the point robots make massive amounts of money. 7000 hours a year times $20 is $140,000. After 1 month, Optimus is paid for. If you bought 10 Optimus bots, you’d make $120,000 a month or almost $1.5M a year. Optimus will change everything. We will be fighting to buy them.
@@johnlehew8192 - Will we, or will humans simply be forced to offer our labour at $1 per hour in order to feed ourselves?
We America, will need to tax the robots per hour. And , Americans will need a minimum universal income.
In a nutshell, yes. For a couple of decades it will need to work toether with highly capable humans so it is no like it will eliminate ALL jobs. It will eliminate some, create others and things will change. For the better or worse is hard to say at this time.
@@johnlehew8192 you're assuming that you will be able to buy one. They may be available by lease only.
Cleanerwatt: Teslabot is worth Trillions $$$
Also Cleanerwatt: “I am NOT invested in Tesla stock or securities”
Me: 🤔
yup, he has zero conviction. Pure clickbait garbage. Laziness just slapping random guy's video onto the channel.
Me: Teslabot is worth billions, maybe trillions. I am invested in Tesla stock for this reason.
@@michaelnurse9089 You should reconsider. There are more advanced bots from other companies.
@@michaelnurse9089 Space is the place where these robots will shine. The limitless resources of the solar system will finally be exploitable.
IMO, non-humanoid robots will ultimately be far more capable and useful than robots in a human form factor. A non-humanoid robot possessing multiple limbs, sensors, etc (i.e. more than a human) that can transform itself to optimize itself to suit the operating environment and tasks it must perform will be of far greater utility/benefit than a humanoid robot.
Problem with this is if you replace all the workers with bots then no one will have an income stream to buy the stuff the bots make... kind of self defeating. space exploration and other dangerous activities makes sense.
The ones pushing for robots are pushing for UBI. The planet will become a nursery.
Th super abundance = virtually free goods and services
@@SparkySho not in human nature to make things for free.... What business owner is going to invest millions in bots so they can make no money and give things away virtually free?
@@andym1548 they’ve been saying this since the beginning of the industrial revolution-so far not true
@@CathyMartin-en2tl tell that to all the car production line workers that lost their jobs to robots. That scenario was 100% about increasing profitability and not paying a work force or having to allow workers rights etc.. Now see if the owners will still do that when there is no increase in profit and tell them they have to contribute to a universal wage for everyone to sit and do nothing....
Every person in the first world will have a bot.
Very improbable because it would imply that everyone, grownup or child, could afford monthly payments for the robot in addition to the cost of having shelter and food.
@@sandmehlig In time there very well may be no cost to have a bot.
As far as demand for robots here on earth. If you could buy an Optimus robot or two or three that could cook and clean and take care of you so that you don't have to move into assisted living, which is REALLY expensive for one you'd enjoy, it would be significantly cheaper (less expensive) to buy an Optimus robot or two or three as assisted living that you wouldn't hate is $5,000 a month which most people can't afford.
Even at $20,000 for an optimum robot would still be a lot cheaper than a "nursing home" where basically the person would be laying in a bed and would need general and mostly easy medical care plus house cleaning and feeding of the person.
Also keep in mind that Optimus robots have 4k cameras in them (or can) and that would allow doctors and nurses to remotely give a good and easy overall examination of a person's whole body REMOTELY. This would make it SIGNIFICANTLY less expensive then hiring someone to get the person to the doctor if it is just for a checkup.
And if the person is semi mobile then the Optimus robot would be programmed to be caring and wouldn't have good and bad days. And it would be strong enough to help the person in any way that they wanted. Plus they wouldn't have certain work hours or families to go to nor would they need sleep.
What about charging? Simple! The Optimus robot could literally plug into a cable which would be long enough to take care of the person in their bedroom and could assist the human in anyway needed as long as it was at bedside. And by the time the person needs/wants to get up, the Optimus robot would be fully charged. it would only take 15 minutes to have a pretty good charge and way less than an hour to get a full charge. But they could plug it at any time of the day or not they aren't needed with a cable that would be long enough for a area of maybe four to five meetings which is about twelve to fifteen feet.
This care would be SIGNIFICANTLY less expensive than any assistant living or even nursing homes until the later part of the person's life where they are very, very near the end of their life. Then an Optimus robot could use a Tesla car to get them to a doctor's visit or to a hospital. Actually I expect that Tesla ambulances will be available at some point specifically for the use of Optimus robots that need to transport a person to a doctor or to a hospital. I think they should start thinking about sooner than that later as within eight years I could see that being another segment that would pay off for Tesla. They would need some kind of electric van for that which wouldn't be hard for them to design and build.
Exactly, Elon thinks these will be selling for 10k when they get manufacturing up to scale. Much less expensive then a nursing home. If it can do everything a human can for a one time cost of 10k. That is the end of nusing homes.
Some good thoughts there, but as far as calling in for a remote doc visit, you are living in the past. AI has already surpassed human docs in diagnostic abilities, and it will be so inter-meshed with you daily activities that you will never have to see a doc unless you need a procedure, and even that will probably be done by a robot during a house call. Your health and condition will be continually monitored by brilliant systems at all levels, thousands of times an hour. You will be immediately notified of any changes that need attention.
What is the cost of electricity these robots need to function per hour? This needs to be considered.
The key to robots will be software / AI. The physical side is done for basic human tasks. Tesla has AI and mass production expertise - I don't see the same in any other company.
Tesla has this but so do other companies - particulary those in China. China is often bad at software though...
Well the explains the intelligence of a toad.
Very nice presentation by both of you!!!
Tesla will probably sell app like software packages for different functions, and focusing on this area would probably be worth more than the bots themselves.
And the bots will stream commercials
Industry will embrace the robots where possible and eliminate jobs when given the opportunity in order to maximize profits. Without jobs, people won't be able to afford to buy robots and universal income will come into play. There needs to be a fine balance, otherwise we will see a even greater divide between rich and poor. US companies gave up manufacturing to China and surely we will give up jobs to humanoid robots but at least we can bring back manufacturing to the US.
As they should you know u would if u could
I bet there ain’t a speculation th TTTesla team ain’t b thought about
Can't wait for an affordable home robot.
Keep waiting, Elmo promises much but delay and delay.
@@DrRussPhdbroke bum hater
The subscription model has a glass ceiling of the minimum wage. And I don't think it will make any sense to replace a human unless the hourly cost for a robot is at least 50% cheaper than hiring a teenager or outsourcing it to Vietnam.
So charging a premium on these robots will be forever out of the equation, and so it makes no economic sense to think that these robots will replace cheap labor in a global market lol
Imagine if Netflix pitched this magnificent business proposal about never raising prices ever, stuck at charging 7 dollars forever ever, who on earth would invest in that? What that financial adviser is saying about eventually raising prices as their capabilities are improved, it is just ridiculous. As a low level cheap labor I need it to do one programmed task well, move this box from here to there, check inventory, scan items. That's it. I don't need more enhancements, and I will not pay for any other upgrades ever, I don't need it to jump or even better cognition. The best paying jobs are managerial, decision making and creative positions, and those don't require physical robots, AI will be humming quietly in a server room for those roles. You don't need a body for that.
Plus, imagine having a whole plant paralyzed for a bad update on all your "workforce", freaking nightmare.
The whole Optimus thing is doomed to fail: for industrial uses, you need specialized machines not human like inefficient and ineffective anatomical copycats rofl.
People will always hire a cheap immigrant to clean the house, not a 30K USD bipedal roomba with less dexterity than a 3 yr old, what are you guys smoking
Why would there not be “apps” for different functionality that can be subscribed to? For example, Joe wants his bot to be a gourmet chef, an artistic gardener, and a math teacher for his children, so he leases his bot from Tesla and then goes to the Tesla App Store to subscribe to these apps. Independent developers could create apps for Tesla bots just like others do for iPhone apps in Apple’s App Store. An electrical contractor leases 20 bots for use in his company and gets 20 Master Residential Electrician App subscriptions. A physician leases bots as pediatric “nurses” and ‘physicians” for his clinic (or more likely, each family subscribes to a “pediatrician” app for their home robot). In this kind of situation most of the money would be in app subscriptions more than the bot itself.
dude stop trolling.
@@wiziek help me understand. Wasn’t trying to be a troll.
Yes. Tesla has an incredible headstart with their robot car fleet that can translate quite easily I would imagine to other incorporations such as humanoid robots
They are not, even Asimo by Honda was more advanced, and that was almost 20 years ago. That Optimus walking is a freaking joke
Before I hit 80, I want full self driving and a home bot.
Soo true, war is often the most important & difficult application!
I appreciate the analysis, but it reminds me of how politicians justify placing a tax on certain activities; that the affect of a tax will not reduce the activity as people and businesses avoid the activity in order to avoid the tax. In the end, the envisioned tax revenue isn't realized because it made the activity too expensive.
Similarly as bots are added to the $16/hr labor market, human laborers will be freed up, causing more competition among laborers.. Some will work for less than $16/hr in order to find work and compete against the robots. Some human $16/hr laborers will upskill to find the $18 and $20/hr, driving those rates down because of the added competition.
Ultimately, I think the cost of labor will significantly be reduced, diving to near $0. These costs outline in this video will slide as the affect of robot automation creates lower cost workers.
The only tax that's needed is a tax on the income made by the robot producers. Why would we tax the bot itself? We currently have millions of robots in factories already, and they aren't taxed. They are considered equipment. Just because a bot can walk shouldn't change the tax structure.
His subscription and sale prices are a complete fail because he is not accounting for competition. If he charges $300k and the Chinese charge $15k I will buy zero from him and three from the Chinese.
The concept of a humanoid robot for industry it is actually stupid. You don't need "generalist" robots, you need highly efficient and specialized robots.
Anthropomorphic robots for business makes no freaking sense, absolute zero sense.
Boston dynamics have very interesting designs with legged-wheeled robots which I can totally see them being useful in a warehouse. Bipedal and humanoid? That's really stupid unless they are used for interaction with humans, like companionship and monitoring of elderly care.
@@xantiom i think you still dont understand why humanoids make sense. You see, when it comes to robots, hardware problems have pretty much already been resolved for the most part, the main issue holding back robots is the software. Boston robots may be able to do a lot of programmed hardcoded fancy moves, but it takes a lot of effort to code it to resolve new kinds of problems. For optimus, the main goal is to resolve the software hurdle. how? By imitating humans via data training, lots and lots of data training, because humans can resolve a lot of very complex problems, you just need to copy and paste and let the robot imitate. Saves way more time than coding done by boston. So if you build a humanoid robot, data is freely available to copy and paste without much need for altering the code. Non humanoids dont have a ready available source of complex data to copy from.
I think Cern Basher is wrong about Tesla Robots being off planet "later" and here's why.
I think that sending humans to mars in quantity is going to be SUPER risky. And keep in mind that if we WANT to do that, we need some place for them to live.
There is a company that has used a regolith substitute to 3D print buildings but you have to get that onto the Moon, you have to assemble it, and you have to take real regolith and supply the 3D printer with that material.
I think it would be SIGNIFICANTLY safer to send Optimus robots to the Moon and operate ALL of the equipment to do EVERYTHING. Literally everything. It would be easy to upload all the different "jobs" that Optimus robots need to do to Starship either before Starship leaves earth or earth orbit or even when in earth orbit and then have Optimus robots "plug-in" like Borgs did in Star Trek (though with less cool looking hardware) to get their programming.
Keep in mind it will be SIGNIFICANTLY less dangerous to get them into space, get them to the Moon, get them ONTO the moon and then have them assemble the 3D printer that will print the buildings for humans. They could also assemble everything they need for themselves from charging stations, to vacuums that would vacuum Moon dust out of their parts, etc., etc., etc.. And these Optimus robots could drive whatever equipment you have. They could even be programs to "operate" on each other as parts get destroyed do to Moon dust which is very corrosive.
Also, most of what exploring can be done SIGNIFICANTLY safer with Optimus and they would be much easier to house since they don't need oxygen or anything else to "breath" and the only thing they need to feed on is electricity. So easier, easier, easier.
And that's just the Moon. I believe the project on Mars where Elon wants over a million people there, which is what you would need to be able to build a city and have enough humans there to protect us from ourselves or some unknown "thing" like a lot worse COVID or a lot worse asteroid than what killed little Dino from the Flintstones.
exactly - Tesla will need to reduce e.g. a Martian city to the absolutely minimum bootstrappable seed that can be loaded onto a fleet of starships and sent to e.g. Mars where it can start with the acquisition and refining of raw materials, then going up a ladder of producing more and more elaborate tools, machines and factories until you have the components you need to build lots more Optimi who can then build a city and do some terraforming before the Humans actually arrive. Except that Humans will probably be involved in the process the first few times until it can finally by shrinkwrapped and scaled at which point we can literally seed the Universe as each seed's job will simply be to produce more seeds (a seed includes a Starship) - just like a tree.
Thank you - you’ve made he think.
The first person on Mars will be a robot.
Tax will need to be paid on each robot.
I started 8 companies over my lifetime
This is a game changer
I would have absolutely saw the opportunity to make this my new business opportunity
Absolutely a no brainer
Power grid will need to be upgraded.
The debate has never been over how valuable an Optimus robot would be if it were as capable as Musk implies it will be. The debate is over how long it will take to get there - months, years, or decades. Full self driving has taken far longer than Musk predicted, and it's still not at level 5 autonomy. The AI problem that needs to be solve for Optimus is far harder than that for FSD.
Who will afford a robot when they take all the jobs? It is a tricky thing and life may change more fundamentally than people think. We may no longer have a use for money. Robots make robots, robots mine the materials deliver the robots, no one needs paying.
I’m getting close to 59 yrs and own 355 Tesla shares (all in). I’m hoping to be able to buy an Optimus bot to help me out in my later years.
I love Cerns analysis
Sounds like either sell or lease for a base price with minimal skills, and allow 'downloadable upgrades' like being done with the Tesla to sell or rent '$20/hr equivalent' or '$40/hr equivalent' skill, all on top of the 'base capabilities' that can do quite a bit.
Even for home use, a base 'cleaning bot' that could also do laundry, and a possible upgrade to a 'cooking' robot, to a 'Mechlin Star Chef' robot skill level might be reasonable, and do the same with manufacturing or construction skills.
Each skill could have a time/lease cost option, so you only need 2 hours of a 'cook' daily, but you might need a weekend Michelin Chef' rate for a big party. And different for feeding 2 or 4 or 400 people in that time.
Will they replace tradesmen? Plumbers, roofers, electricians?
It’s important to bear in mind that Elon is primarily a physicist and thinks in magnitudes. When he predicts 10 billion robots what he really means is that he’d be surprised if it was a magnitude less or more, so he’s effectively giving a probable range of 5-50 billion.
$250 trillion market cap company? There is only about ~$130 trillion collectively on the entire planet.
Fortunately market capitalization is kind of an imaginary thing that doesnt have to be backed by actual money.
@@thirdplace3973
Yup. Yes - this is what happens if you generate 15 billion more units that can do human labor in 5-10 years. Normally this sort of thing would take hundreds of years.
@@JohnBoen But where is the money coming from to put into the company for it to be worth $250T market cap?
The planet certainly didn't start off at 130 trillion. It increases according to its productivity
@@LowkeyXxx it increases also because of fiat currency and printing money out of thin air. Minimum wage used to be $1.25 per hour. 5 solid silver quarters. Today that would be worth $30/hr.
1. Humanoid robots might raise human and industrial powers but only by overshooting the efficient consumption of resources in service of the multiplication of human productivity.
2. The efficient consumption of resources has regulated the operation of industrial firms and workplaces broadly throughout the whole history of industrial capitalism. (Efficiency gains come in concrete steps however so ineffiencies can persist even as newer and more efficient production techniques supersede older more problematic techniques.) Capitalisation serves the goal of lifting productive powers and reducing costs of outputs when those outputs are produced in volume. Surpluses grow because despite increased capital spending the cost of production per unit of output still gets reduced/economised. Similarly, as surpluses/profits increase these can be recirculated into further capital expansions so long as this can be done efficiently.
3. Only efficient production (usually implemented in a phase of capital expansion and with a pay-off over the long term) gives one firm a practical edge over another. If humanoid robots don't lift production - they don't - they are utterly useless. Currently, the advocates of humanoid robots are promoting the bug that they don't raise productive powers (at a competitive capex investment) as if it were a feature. Companies that go down that path will quickly find themselves in financial trouble. We were already on the right path with robotised machinery and tools that increase the speed of production and reduce per unit costs. This humanoid robot thing is a complete non-starter for industry.
4. Non-humanoid robots multiply human powers efficiently because they increase the speed and efficiency (reduction of resource requirements) of production. Humanoid robots don't do that and any business that spends its profits on expanding uncompetitive/inefficient/overpriced productive output will fail.
5. Humanoid robots only make financial sense as a replacement of human workers without any compensating UBI. And, that is a crap idea.
The uncertainty I have is about how intelligent Optimus will be. That's how easily it can learn new tasks.
Simple repetitive tasks will be enough at the beginning. But more flexible ones, like household work, is possibly out of reach for a long time.
This can easily cut the profitability by a factor of 10.
If a Tesla with FSD can drive itself then I think an Optimus, with software many generations on from the car, will be useful. Many people all want the same domestic tasks, so once an Optimus can safely navigate around domestic settings it won't be long before each s/w release (monthly, weekly ?) brings you new skills, just like is currently occurring with the cars (c.f. recent release of Summon capability) :
- vacuuming and cleaning
- washing and ironing
- cooking
- mowing and gardening
- general DIY
- more specific DIY - carpentry, plumbing, electrics, heating...
- ...
And this is assuming s/w releases being structured like Tesla cars i.e. in serial but in order to scale, development will have to become modular and proceed in parallel like an AppStore with multiple companies producing training data in their particular area of expertise ... In fact you will probably have a whole selection of e.g. Cooking apps just like you have a selection of different restaurant types - Indian, Chinese, Thai, French etc.. It is going to be an amazing time to be alive !
You can't assume that Tesla is going to corner the market for those robots. It seems to me that Unitree and Figure 02 are already superior to Optimus. It remains to be seen how they will compare in usefulness or price. Optimus will be one of many available options.
The central question remains: what happens to displaced humans. Watch Elon and Sam Altman discuss this a few months ago and agreeing that first to go would be blue collar jobs, then white collar, then creatives. Software developers are already getting squeezed by AI. This guy shows no concern or appreciation of the depth of the problem. Talks about replacing all labor. Not a single word of concern, mindless drivel about trillions of dollars.
A traditional answer is that new jobs will emerge - but these would be also taken by robots. Humans would have no jobs if robots are so capable.
I think the answer is that post-robotic humans won't have "jobs" and certainly won't "need" them as essentially for survival. What humans have and still might have are "responsibilities, duties, goals, desires, etc." which will fully occupy their times. In other words, humans "will get a life" in trade of "jobs" and "working for the man".
In the past, old manual labor jobs were replaced by new manual labor jobs. If all manual labor jobs are filled by bots, then UBI is the only solution.
@@darylfoster7944
Not really. There are lots of what used to be manual labour jobs that are now accomplished by machines. Milking cows and telephone operators are an example. There are countless more.
I think the lease vs. buy argument is wrong, I suspect it will be more along the lines of high end CAD systems, you have to buy the robot outright, then you have to pay a monthly maintenance fee. They might require the maintenance fee also, not make it optional SO you get both the sale and the monthly fee.
But will the healthcare insurance industry allow their costs to be covered as medical needs?
I read that by 2035, Tesla could produce around 200,000 Optimus robots annually. So how do they get to 8 billion robots anytime soon?
However you have to understand the gross number on human robots is the total market, but Tesla aren’t the only ones building human bots, so Tesla’s potential value will depend on their market share.
The main problem here is this is untraveled Territory, guesstimate are subject to large errors!
First step is energy. Once we have the "super power" we need for the growth curve of bots it will happen. My rough estimate is seven years.
@9:39 You don't understand the products. There's the robot(s) and then there are the skill(s.) One robot can master many skills.
First you lease (or sell,) the robots, and then you lease (or sell,) the skill(s), and
then you lease (or sell,) the equipment to train the robots in new skills and then you lease (or sell) the ability to do OTA updates to specific robot units.
how many materials and batteries would be needed to build 30 billion robots?
Raw materials is 100th of a car - 30kg vs 3 tons.
@@michaelnurse9089 so lest say it is a little more, about 60 kg. it would mean that tesla woul need the raw materials of 600 million cars
The calculation doesn't work that way because there is cost pressure from competing products. In the end, such a bot must and will be much cheaper and this will promote its adoption. Competing products will probably be more successful because they are already more developed than Optimus.
"Everyone on earth will want one. . . ." -- Elon
"The slave does not dream of freedom. They dream of having a slave of their own." -- Cicero
I actually don't know if they will ever go to a sales model.
The brains will constantly improve, and ultimately, they are selling capability.
Probably easier and more profitable to lease them out... and safer.
I can think of other reasons to stick with a rental model:
1) Waste/Recycling control:
It would be bad if old robots weren't disposed of properly, from optics to environmental to regulatory.
2) Unauthorized Repurposing:
Could you imagine if these things were turned into some warlord's personal army, or into gangland enforcers.
3) Competing with old models:
Why give customers the choice to save by sticking with old stuff?
4) Data Collection:
You want as much as you can get. It's the life blood of an A.I. company.
5) Control:
Why give any up?
These are just a few things I can think of that say a sales model is bad.
Maybe competition will force your hand and you'll have to sell, I can't say for sure.
I just know I'd stick with a subscription model for as long as it was viable.
Most of the humanoid robot labour market doesn't yet exist. Once they are made en masse and work well, there will be more jobs for them that humans DON'T yet do than currently existing labour.
The majority of people are having trouble with groceries and gas, $20,000 for a robot to buy or rent! What a scam😅
5:09 Imagine how much effort it would take just to build 1 billion robots alone. Tesla stock would have split 5-6 times by then i’m sure. Maybe 10 times once they ramp up
Have no idea when these robots will be done. 10 years, 5 years, 3 years Can the robots run into fires and carry people out? Can they move as quick as people do
Optimus will be used in Tesla factories this year
No they can’t move that fast YET….
Terminator could walk through fire.
Trial production late 2026 - full production 2030 - growth stops 2040.
Unitree has their prototype going 6 mps. I think if you want to use them for fire rescue they must change the form factor from human to crab - walk up the side of the building and carry the fat lady back down.
What prevents powerful countries from eventualky weaponising humanoid robots?! ☠️
Yes world wide there may be 14 billion phones in use, but the majority of those are low cost phones. Unless the price for owning or renting/leasing a robot comes down to a few thousand dollars selling price or 100-200 dollars per month for rent/lease, I don't see a majority of the population having a personal or even family robot. I can tell you now, that my income would not be enough to even be able to afford those prices. And having robots take over a lot of production jobs, would also mean that for a large portion of the workforce, their jobs would be on the line, meaning no income and no way to afford a robot. Unless of course a form of a decent universal income would be set up and I don't see that happening in the current capital doctrine that runs the western world.
Sadly I feel that the way society is run at the moment, robots will take over more and more of the lower skilled labour. And with every increase in capability, the next tier of jobs will cease to exist. As a result those people will lose their ability to buy stuff and keep the market going, unless that universal income is sorted. If not, be ready for some serious, and I mean French revolution serious type of unrest because people at the lower end of the economy will have no choice but to fight for their survival.
In short, I am not sure, those that try and increase the use of AI and robotics have thought through how disrupting this will become for society.
ya i burst out laughing... oh there's 14 billion phones ... therefore theres gunna be 14 billion or more bots! Real high end quality logical analysis hahaha.
The cleaning company can have a fleet of them. You effectively rent one for 4 hours a week.
Would you rather be taxed to pay for space development or just buy really cool products and services?
Many comments say can't quite replace human labor..yet,
Though the robots to Mars to build shelters and greenhouses and Boring company tunnels, is outrageously expensive and probably happening soon
Guaranteed to break after 12 months. Repair bills will mean a junk yard full of them just like a Star Wars movie
You never heard of a guarantee not to break after 5 years?
I’m very skeptical of any savings being felt by the consumer on this technological Takeover
Something has to be done about Worldwide inequality - if not you will be correct and all hell will break loose
A functional C3PO would be nice to have. Robust automatic speech recognition remains elusive (1-8 meters lips-to-mic, spontaneous speech, multi-talker, etc.) Entropy is also a major hurdle for current AI approaches. A 800 watt hardware-5 cpu might work in a Tesla car but impractical for a functional 3CPO. First princples are being ignored. Vaporware for now.
Imagine a pizzaria with robot delivery service on wheels?
If it cant cook and clean and barely walk or figure out its environment its a pipe dream
A thing about market is when labor market is replaced by robots then what replases market when wages are not used to consume goods.
Other robot corporations. See Blade Runner, or Star Wars, or virtually any other scifi. We will be to the robots what dogs are to humans.
Imagine these things in the army - no recruitment problems. China will be right on it.
We need robots that remove trash from the ocean
Bots' earnings will be taxed as human labour currently is - income tax.
Maybe somebody should tel Elon that half the world can barley feed itself and 20% dont have reliable water or electricity.
Then put those robot to work.
Food is a distribution problem not a production problem. We harvest more than enough food to feed everybody 3 meals a day. The problem is that the distribution is uneven. More food is delivered to you than the amount delivered to the starving children in Africa.
We hope that Robots will make products plentiful and cheap.
Most of the World is fat today. Too much typing in the RUclips comments section.The irony runs thick here today.
@@michaelnurse9089 💯
A world without Human Recourses Officers 🍻🎉
It seems people mostly seem to think of robots (humanoid, drone, wheeled, whatever form factor) as smart machines. Maybe, your Tesla is and may continue to be just that, even when you can talk to it, like Siri or Alexa. But actually that changes if AI approaches sentience and consciousness. The situation also changes when AI robots are brought into the home or the small business workplace. At that point, they will cease to be like dolls and teddy bears, washing machines and vacuum cleaners, or computers and phones full of confidential data. They will become part of families or business teams, purchased initially maybe but not leased. They can be enhanced or up-lifted, but not switched off or erased or wiped when updated. Domestic applications will need stable personalities. In the meanwhile, smart machines can do dull, dangerous, repetitive work at levels of reliability, quality, and speed far exceeding what humans can do and historically, up to now, have had to do for themselves.
You're crazy to think if anyone's going to pay that exorbitant amount of money when you can buy a Chinese bought eventually for $5,000 so the market will force the sale of the robots cheap
Just imagine the Humanoid robots doing separate tasks and the recurring revenue it can generate?
#Massage therapist ,
#Chef
#Cleaning lady
#Gardener
#caregiver
All these could be separate apps downloaded for each tasks.
You missed friends with benefits.
There are many other companies in the world who compete with Tesla. The price cannot be dominated by Tesla only
It's not about hardware - it's about software - no-one currently competes with Tesla on software - Tesla is the only company with millions of [almost] autonomous robot [car]s working out in the wild... - oh and they're pretty good at hardware too.
The main problem with your analysis is it lacks the appropriate inclusion of competition in robot production, programming/training, etc. For example, the Ag industry is now almost fully automated, increasing the human productivity by more 100-fold. The same will be true for all the other areas where human labor's productivity can be effectively multiplied more than 100-fold utilizing robots. Competition will be the primary constraint on the ability to profitably deploy any particular format or specific type of robot, as it should be. Regardless, there will be a huge first-mover advantage in the rapidly emerging marketplace for these super-flexible, human labor multiplying robots.
Any new “way to “ change the way the world operates or gets anew
Like cars
Cell phones
Etc
The robot new world
Will be a game changer
The robot market will be huge but there will be plenty of competition when the market develops. Tesla won't be the only company selling them.
One gallon of gasoline is approximately equivalent to about 25-30 hours of human labor. Human work 35-40 hours a week. A gallon cost 3.50-4 $ per gallon in USA.
So a human physical labor production in gasoline price equivalent to a maximum of 8$/week ( 2 gallons) but cost 800 $/week at 20 $ / hours. 100 to one.
Robot cost 20 000 $/ 800$ per week = 25 weeks to pay back the robot. The robot can work 2.5 times longer then a human ( 20 hours/week). So the robot pay itself in 10 weeks minus operating cost.
I think this 20 hours a day thing is wrong - there is no way I would let an expensive piece of hardware sit uselessly for 1/6th of its time. Recharging will be done as it does its job, either wirelessly by standing on a charging mat if the job is stationary or on a cable if moving around a bit. There will be very few jobs where this cannot be managed.
Peeps don't seem to understand that the bots work for free.
Musk is absolutely brilliant in certain ways. One example is his epic ability to keep the Tesla stock price elevated via a series of "hot new technologies" with "immense market potential" which keeps his devoted followers buying. The smarter traders have left long ago, recognizing what lies ahead. The bot concept is interesting but a close look at what it will actually take to get to markets is sobering. Regulatory approvals for widespread applications is likely a decade away. Safety concerns in human-bot environment are infinitely and regulators know it. But these realities are not Teslas most immediate threat. It's musk's imminent mental breakdown that will bring the stock price down to earth.
What military applications does Optimus have?
Tax system will need to see some wild changes.
If you assume an average of 2 Optimus robots per human and Tesla charges just $1/hour, that's 16 billion * $1/h * 24h * 365d = $140.16 trillion per year of revenue. And that assumes they charge nothing for the robot. Even if you made MUCH more conservative assumptions, no matter how you look at it, you're talking about some truly wacky numbers.
I believe he said one robot could replace 3 factory workers by working 20 hrs a day. That's all three shifts.
So does this theoretically mean we may see massive deflation?
high-tech has always been very deflationary. Look at what a PC with a 386 CPU cost back in the day. And what a fraction of that money can buy today.
Yes
I've been to many poor third world countries. Even in these poor countries they still all have mobile smart phones. These people will NOT all have $25000 personal humanoid robots.
How's the Tesla self driving car going .... not too well
These valuations are egregiously over inflated. If the bot takes $15k to make, we should expect competition to bring the cost down to cost + and acceptable margin, e.g Apples extremely healthy 30%. Thus not a $800k bot, but rather a $20k bot.
The cost to own one may grow once governments start to look for income tax revenue replacement.
Since industry can use bots to make bots, the smart company should be able to scale exponentially, so I don’t see a supply shortage lasting very long - as long as the parts themselves are not constrained.
"Since industry can use bots to make bots, the smart company should be able to scale exponentially, " Only in theory. You need machines, raw materials, management, capital and so on - nothing has ever scaled exponentially for more than a short period (outside the spacetime expansion of the universe).
He didn’t take into account productivity. These robots are slow because of hardware limitations and power constraints, regardless of how good the AI is. A robot working 20 hour per day will only be able to create the same number of widgets as a human can in 10 hours. The value will be more like $1,300 per month per leased robot.
Where’s the power constraint? You realise the bot could stand on a wireless charger or plug itself in right?
14 Billion phones? How is that possible with 8.5 billion people on earth
All those illegals pouring across the boarder have 10 phones each.... lol
We will be fighting each other to buy these when they come out. 10 bots can gen $1.5M in labor a year if people pay current labor prices
COmpetition will bring profits of bot owners/users down. Bot makers on the other hand will make trillions...
When they anounced Optimus they were kind of the first and special. Now we already have 15 very hard competitors that seem even to have better bots already now. There will be so many bot companies, the prices and margins will ve very low. Think about it.
its about the software not the hardware ...
Awesome guest, but what does he have against CERN?
So, I wonder how much a PHD is worth to mow my lawn
I dunno, the big assumption is that a humanoid will eventually be able to do everything an average adult can and I just don’t see that happening. AI-assisted humans? Certainly. Dangerous, boring, repetitive jobs, of course. But I still want a real person (who has a soul like me) to cut my hair, to teach my kids, to make my food, to take care of my elderly parents, to conduct my physical therapy, to clean my house, to coach me through my career development, to pull a splinter out of my finger, to motivate me in the gym, to hang out with. Crap jobs will be automated but there are many more things that only other humans can do for the best results.
As a geek and Tesla shareholder I am excited about FSD's future.
Is Elon mistaken when he says that Tesla has an equivalent moat when it comes to applying FSD to Optimus?
Tesla doesn't have a fleet of robots being remotely controlled each day like their cars...
My concerns are specific to the numerical disadvantage of employees vs a huge fleet, as well as the fact that whilst the driving world is infinitely variable, the goals are highly defined. A general purpose robot faces goals which are proposed to be infinite as well.
By my reckoning, this leads to a situation where Tesla starts much closer, albeit ahead of the competition, in a race to an outcome which is much further away, and probably much more limited that the proponents' expectations.
As an outsider to the field, I am far from certain of my take and would value any input both from intellectual curiosity and for my due diligence re my Tesla shares.
Also a shareholder. FSD has a massive moat - 7yrs+ head start. Optimus has no moat. It will have to rely on Tesla's access to better engineers, more capital, and Elon's instincts. I would stay invested for both. Upside oportunity is high and downside risk low.
Factor in all the more advanced robots which are already coming to market like the Electric Atlas from Boston Dynamics (Hyundai). This video seems to gloss over Tesla's bot competition.