How Much Would You Pay For a General Purpose Humanoid Robot?
Humanoid Robots will change humanity forever
Optimus, also known as Tesla Bot - Source
This article is primarily focused on residential needs in developed economies.
Over the last year or so I have been asking all my friends, mostly in the United States: How much would they be willing to pay for a humanoid Robot? A humanoid Robot that can do practically all your chores: clean, cook, drive, medical care, like you never have to hire a person ever again, instead you pay for an app to enable any given skill.
While my research is not statistically significant, I’d say, most were very comfortable at $10K, some $20K. Don’t think anyone went beyond that.
I also ran a poll online across all my social media platforms and while I got a tiny response (not statistically significant), 60% said they’d be willing to shell out $30K for a robot like this. 20% each for $15K and $100K+.
So anecdotally, I would put the results of the survey at $25K.
My take is that one would be willing to pay as much as they do for a car. I think its way more valuable than a car, in fact your car does not need to be that smart, if you have a humanoid robot. Extrapolate that for a moment.
Speed of Technology
Technology is moving faster than we ever imagined. Heck we’ve had digital technology for only ~50 years or so. And the speed of evolution is exponential.
Think about what we’ve done in the last 10 years vs the 10 years prior to that. Remember 2003? Did you have a mobile phone then? Which phone was it? How does it compare to the phone you had in 2013 and then now? This is when phones are not even at the cutting edge of technology today (which could be changing).
Humanoid Robots will change humanity forever. It's as if, as a civilization, we have been working towards this goal for as long as humans have existed.
We first invented simple machines which went through 1000s of years of evolution, we used mechanical tools to make tasks easier for mankind. Then we built more complex machines which did not require human muscle to move, increasing productivity many times over.
Then we built the digital world, where we can do more complex things faster, quicker, easier. We gave the digital world the ability to talk to all digital life across the world in an instance (almost). So everything is connected.
Then we built a synthetic brain, a neural network which builds reasoning, ability to do multiple tasks, and learn new tasks using machine learning leading us to AI.
And now we are closer than ever to build all that machinery, connectivity, reasoning into robotics.
V1 of that is your Roomba & V2 is likely going to be your self driving car. But both those examples are single purpose with limited reasoning and ability to learn.
V3, however, a humanoid robot is a totally different zip code.
Listen to Balaji explain it using the idea that we have moved from mining muscle out of the ground to using silicon to create an artificial brain - Clip.
“Replace muscle with coal and brain with silicon” - Balaji - Source: BeerBicepes YouTube
This AI revolution along with robotics is as powerful as the Industrial Revolution and I would argue it is several times more powerful, it is truly exponential and I don’t think we can comprehend the scale of change it will bring to our life in the next 1 to 2 decades. I think it will be an order of magnitude higher than what the industrial age did for our ancestors.
Listen to Balaji share how he thinks this is on the same scale as Oil - Source: BeerBicepes YouTube
Sidebar on Nvidia
I would also like to take a moment and acknowledge Nvidia here. AI will eventually be a bubble at some point, we are not there yet. We are in the 1st innings of AI, Nvidia is the arm’s dealer and the only one at the moment. Even if serious competition does show up, we don’t know when it will hit high volume manufacturing and how it will compete with Nvidia’s chip. Also Nvidia is not sitting and waiting for competition to catch up.
Nvidia and Tesla are the two most important companies of our generation
Nvidia and Tesla are the two most important companies of our generation. We are re-building the world and Nvidia and Tesla are going to be integral parts of it, as integral (if not more) as Google is to the internet today.
I would also go to the extent to say that both Nvidia and Tesla are part of our new big tech leadership. If America’s focus is indeed moving back to manufacturing and the real economy, then these two companies are at the front of the line. (Though Nvidia is completely dependent on TSM to manufacture these chips and Tesla is also heavily reliant on China mostly from a revenue point of view, unlike Nvidia).
The Tech:
I am going to simplify this. Oversimplify, in fact.
A humanoid robot needs to be able to understand instructions (which is math and code) and then implement what is asked for by moving around on its feet and using its hands to hold/move things. I know there is a lot more complexity to it which I am not qualified to understand, leave alone explain it. But if a robot can do basic movements like this, I think it can be very useful with just that.
Here is another clip from the same podcast, where Balaji explains the 6 degrees of freedom arm.
I am no scientist but a robot made me coffee in San Francisco and it could very easily hold things and move things around. And that coffee machine has been there for like 5 years now.
Agreed, It is not general purpose and it can’t move to change its location, so it is a far cry from a humanoid robot.
However, this is not!
What Could Robots Do For Us?
What are the most common things that a robot is likely to do? Everything we don’t want to do. All the chores: Loading and unloading the dishwasher, loading and unloading the washing machine and organizing the wardrobe, cleaning the house, dusting the carpets, sweeping the floors, organizing, taking out trash, growing fresh vegetables in the backyard and keep you front yard looking great, always as well keep you home safe and so on.
Some of these things a robot is capable of doing right now. However with a general purpose humanoid robot, with the added ability to move, hold things with its hands and sort them, these tasks are the easiest for a humanoid robot to do. They are relatively low risk too.
The next thing I think is essential is the ability to cook. Cooking is, for the most part, a science. You have instructions, which when followed correctly, should always give great results. Remember a robot's ability to know exactly what temperature the gas is on and its ability to have impeccable timing is unmatched in humans. So as long as it can read and follow instructions, pick and sort, use its arms for more complex motions like more cooking motions like cutting, preparing, stirring, mixing, etc…it should be able to make you great food and it would not be limited to any cuisine because you should be able to download it, matrix style.
In addition to the coffee machine I mentioned earlier, Chipotle is testing robots to make bowls and robots that can cook already exist (and that video is 7 years old). However these are still V2 robots based on my definition above.
Guess what? You don’t have to get the groceries either, the robot either orders it, collects it and stores it in the refrigerator or it can go to the market to grab it for you and bring it back home and cook you a great meal and you don't have to leave your couch.
Self driving is already something that we are very close to cracking, so I don’t need to make a case for why this is helpful. Your car does not need to be smart because your humanoid robot can drive it and it can drive any other car you may have. In all honesty though, humanoid robots will need the assistance of all the extra features a self-car driving car has for it to operate safely to begin with. However as humanoid robots become smarter and their ‘senses’ become closer to what we can do with our 5 senses, then it could be just like getting into a car and driving it like a human, no assistance needed.
The last example I want to talk about is patient care and being around children. It is one thing to have a humanoid robot in the house doing chores, another completely to have a robot be the sole caretaker for your elderly parents or your young children. I don’t think managing children is a primary use case for robots even though I think they can be great at it and can be programmed to be extra safe with multiple fail safe features. These robots can also help educate, tutor your children based on presets that parents can control.
But for elderly care, it would be a perfect fit. Older people are likely to have higher affordability and would be willing to pay for something (someone?) who is always there to help them, always there for companionship and just so much more. These robots can become doctors overnight and are ideal for patient care or elderly care.
My focus for this article has been primarily on domestic or household use cases. I am not even getting into commercial and enterprise but just to mention some of the categories.
Robots are already our primary way we explore space, underwater exploration, medical surgical assistance ($ISRG), etc… Add to that manufacturing, Agriculture, military, construction etc…
Amazon recently announced it will make use of AI-equipped sortation machines and robotic arms, named Sequoia (robotics element).
So again, I don’t think we comprehend the scale of impact this will have. Which is why I keep bringing you back to a more fundamental level of understanding how grand this is, its artificial mind (akin to oil). An artificial mind that is relatively a lot cheaper to deploy once developed, the productivity it will bring is going to be a new age of abundance than mankind has never known.
Robots can also entertain you.
A Robot For Everyone
I don’t think we will be using phones in 20 years from now. With Apple already making a move to a different interface with the Vision Pro, how we interface with the internet and hence the world will go through dramatic change. But I want to compare having a robot to having a phone, something far more personalized than a car. I don’t see this happening right away as these robots are going to be expensive at first but as time passes, I can see every human having their own, personalized robot just like we have a phone, not 1:1 but the gap could be closer than you think.
Within Our Lifetime, Everybody Will Have Their Own Robot - Balaji - Source: Clip
He literally sights Tesla and compares it to a car like I did.
(Most of this article was written on Sept 30th, when I got frustrated debating TSLA 0.00%↑ with a friend who is bearish Tesla but still holds $TSLA stock lol.
The Balaji video came out a week ago. I have used clips from the conversation to substantiate my article but the ideas are my own. I never take credit for things I don’t do.)
Tesla
Amongst publicly traded companies, I believe Tesla is the closest to giving us a fully functional humanoid robot for home usage. Boston Dynamics is doing exceptional work but they seem to be more focused on the enterprise side and I am not focusing on them because we can’t invest in it. (I think Google made a huge blunder selling Boston Dynamics, I will write more on it in one of my future articles on Google)
Tesla is the frontrunner in terms of AI, in terms of having physical prototypes and its ability to manufacture in America at scale and still drive high margins.
I think Tesla is the best publicly traded stock an average investor can buy today, (apart from Nivida) to make a bet on humanoid robots.
How much Would One Pay For a Humanoid Robot?
Eventually, there will be multiple home uses for a humanoid robot, and just like cars, they will be offered by multiple companies at multiple price points.
As mentioned earlier, I think a family will be very comfortable spending the amount they spend on a car to get a humanoid robot. So if you spend $100K on a car, you might be comfortable spending the same on a top end humanoid robot but if you have a $10K car, you might be open to spending only that much on a less fancy humanoid robot.
It’s basically the worth of your time. How much do you value your time?
How much could Tesla make from Humanoid Robots?
Let’s do some basic math.
The US Census reports approximately 121 Million households in the U.S. - Source
Forbes reports 91.7% (110 Million) households in the U.S. have at least one car - Source
In Jan 2023, Kelly Blue Book reported average price of a car in the U.S. to be $49.5K - Source
A humanoid robot is a luxury item, not a necessity (not yet). So let’s say only 50% of all households that have a car, buy a humanoid robot and say Tesla has a 33% market share. That’s approximately 18.15 Million households and with a $49.5K average price, that’s approximately $900 Billion in revenue for Tesla.
Those numbers are based on very light estimates:
55 Million Households TAM, 18.15 Million Tesla’s market share
Does not take into account repeat buys
Does not consider any more than one robot in 1 household
These numbers are US only
Subscription revenue not included
I think 33% market share is light because as things stand now, this market is developing like a duopoly (even though it is very early days). While there have been other humanoid robots, they have not demonstrated such physical capabilities as the Tesla and Boston Dynamics robots have.
Subscription:
It goes without saying that this would be the ultimate subscription business. You want to learn how to paint in Monet’s style, sure your robot can teach you. You feel like having a dish from some tiny region of New Zealand, sure your robot can do it in minutes. Your robot will get smarter everyday and just like Apple's iPhones, the robot or the device is just the gateway, what is really really valuable is that subscription underneath.
Subscription is the real muscle behind this business, the $900 Billion in hardware revenue will be surpassed by subscription at some point. And just like in Apple’s case, the subscription business is likely to have higher margin leading to higher profitability and free cash flow.
Tesla’s Valuation:
I’ve written about Tesla in the past and made the case that I am investing in Tesla’s management team because I don’t know what new products they will release tomorrow or what new markets they will get into which you and I are not thinking about today. Humanoid robots are one example of that.
Since I wrote that article, I have also learned that P/E ratio is the worst metric to measure a growth stock on. Nvidia is the best example I can think of where the P/E got decimated because earnings went up.. I am not saying that I don’t look at P/E or that it is not important. It is one of many metrics you should be looking at and you need to understand why the P/E is what it is and where it is going.
Tesla is an Energy, AI and Robotics company. Take a lot at this page on their website and explain to me again how it is not a tech company.
Tesla’s Margin:
Tesla sold off after it reported lower margins than expected in its most recent earnings report. I mean obviously its margin profile is not looking great.
Tesla is ramping up manufacturing and factory build ups, which takes money.
Tesla is spending billions in AI chips to train its Dojo Supercomputer, which will bear fruits in the future. All this investment is going into the robotics hardware and subscription business and it is being built literally right now and should pay off in the future if Tesla indeed succeeds in building a general purpose humanoid robot.
And the biggest reason why Tesla’s margin is taking a hit, this is the worst possible environment to buy a car if you need to get a mortgage, the only sector that is worse off is housing (which does not have supply, limiting downside in prices, making cars arguably worse than housing). So no shit the prices are falling and margins are falling. Tesla is still profitable and still the leader in EVs.
Right now margins are low for good reasons which will either go away in the future (like interest rates) or start paying for themselves (Tesla Self-Driving, Robotics Hardware and Robotics Subscription revenue) in the future.
Are you going to sit and complain about lower margins today or do you see where the puck is going?
Positioning:
I am at 62% to target on an equity basis on $TSLA. I continue to dollar cost average (DCA) into $TSLA below $250 with small amounts daily.
$TSLA is at support right now at around $215 (based on 1 year and 5 year charts), I will do some heavy buying here today and increase DCA amount and will continue to increase the intensity of buying if the stock falls further. If $TSLA can’t hold $215, then the lows from summer 2023 will be back in play around $150.
I will start taking profits once I cross 100%. No matter what the stock does or what the prospects are. I am here to make money, not to hold Tesla stock.
However, as my net worth grows, my allocation dollar amount will go up by default even if my allocation % remains the same.
I intend to diligently maintain my % allocation until the business fundamentals change or until I am not bullish on the stock anymore.
Borrowed Conviction Rarely Works
Past performance is no guarantee of future results.
The ideas discussed in this article should not be constituted as investment advice.
I reserve the right to change my mind if the facts change.
Disclosure: We own positions in some/all of the tickers mentioned in this article.