JasonDoiy
Investment Thesis
The Tesla (NASDAQ:TSLA) Optimus robot was received to mixed reviews. However, I would argue the point isn’t so much what Tesla has shown, rather than that it is about the potential of how robots could change the whole economy, and by extension civilization.
In other words, I would liken it to the early days of Tesla, when EVs were still a niche, due to their high cost only for the rich. Similarly, while this initial version of the robot may only excel in a few use cases, the question does remain what will happen when it will be able to take over all human labor.
It is bullish that Tesla is leading this new revolution, although it still the early days.
Automating all human labor
One of the primary criticisms when Elon Musk showed the Optimus robot was that it did not necessarily do anything that others, such as Boston Dynamics, hadn’t done before. Tesla was/is seen as still being far behind.
This is, partially, correct. What is correct is that robotics has indeed existed long before Tesla entered this market. However, what is different this time is that the potential for what robots could, in principle, do has vastly increased over the last decade due to the AI revolution.
Over the last decade, as sheer computational power from initially GPUs and later dedicated circuits has been used to optimize deep neural networks, AI has been able to approach and, in many cases, far surpass human capabilities. AI is currently used in a long list of tasks, ranging from image recognition and processing, recommendation engines, translation and transcription, autonomous driving, and more. The latest hype is AI generating images or even movies from a linguistic description.
The point is that all these examples belong to the domain of software, not hardware/robotics, and which is powered by silicon chips. As some people have argued, just merely having a robot that looks like a human isn’t enough, it also needs to be able to do the things that a human can. Well, given the rise of AI, this is now becoming within the realm of possibility.
As case in point, Tesla said that Optimus is powered by the same software as its autonomous driving functionality in its cars, for which Tesla has developed its own chips, both inside the vehicle as well as the supercomputer (Dojo) that is used for training the models. Certainly, these deep learning capabilities (both hardware and software), which are so crucial for making humanoid robots useful, are outside the domain of all other robotics companies. This is where the appeal and differentiation of Tesla entering this market arises.
In summary, while traditional robotics may already have created non-biological technology to replicate humans in appearance, it is only the more recent development of AI that can give those robots the intelligence to do things humans can do.
As one more remark, some may wonder why it needs to look like a human. The simplest answer I’ve seen is the observation that the whole civilization is built around humans. For example, in terms of mobility, a humanoid in the future might transport itself by using an (autonomous) vehicle.
Favorable economics
This – justifying the economics of a robot – is by far the easiest part. The difficult part is convincing, as the previous section did, that it is actually possible in the first place to create a humanoid machine that can do most, if not all, of the things that humans can do.
After all, for thousands of years, the main school of thought seems largely to have been that humans have some sort of immaterial aspect (“soul”) that cannot be replicated by technology. For example, many people remain skeptical about autonomous driving, despite that vehicles from multiple companies have collectively already driven many millions of miles autonomously. So yet again, AI is quickly proving otherwise.
In any case, turning to the financial aspect of labor, just consider the sheer investment it takes to train someone to participate in the labor market. For plenty jobs, people start their career only in their early 20s.
This is in stark contrast to the alternative that Tesla is developing. Once a robot is assembled, which would likely take in the order of weeks at most, endowing the robot with intelligence is just a matter of installing (and updating) the software, which takes even shorter. Once the software is production-ready, installing it is no more work than ctrl+c/ctrl+v.
This is where software has a huge advantage, as all learnings can be retained. In stark contrast, each new human being needs to be trained from the ground up again in all sorts of things like mathematics, literacy, etc. From an economic point of view, clearly silicon + software offers the more efficient approach to create a labor-ready workforce.
Or as Elon Musk said, it would be possible to create millions of these robots at volume production.
And this is where these robots will truly begin to shine, as robots do not need to be paid wages. Hence, as soon as an employer finds out that a robot can do some job just as well as his employees, the economics will simply always favor the robot instead of the human. Especially if it will be as cheap as Elon Musk claimed, at less than $20k. Assuming there is a 1:1 replacement of a human, the robot could therefore pay itself back in less than a year already.
Futuristic implications
The following is just a contemplation of the end-game of what this development could lead to.
Ultimately, it goes back to what economics is about: economics exists because there is always scarcity to some degree (resources, human labor/time, etc.). Capitalism is a way of assigning universal and quantifiable value to these things. Of course, the creation of wealth in the last several hundred years has been because technology has been able to increase the productivity, leading to more resources, leading to less scarcity.
From that view, robots potentially fully taking over all human labor will completely eradicate scarcity. There will truly be abundance. After all, if more robots are needed, it will be robots, all that way down, that will operate the factories – and the whole supply chain – to create more robots. Robots further only need electricity (and batteries) to operate, which is already cheap and abundant. In other words, robots are ultimately just a means of transferring electricity into (for humans) useful actions.
In fact, once there is abundance, the whole notion of money starts to become completely obsolete. There will be no distinction between rich and poor anymore because everyone could have as many robots working for them as required. Since everything is handled by robots, humans do not need money anymore (not even to buy stocks, or anything else).
For example, if you need to do groceries (something plenty are complaining about in current times given the inflation), you just pick whatever you need (or you assign a robot to go shopping), and the robot workforce will simply refill the shelves.
Welcome in utopia.
Timeline
Of course, it is impossible to predict how long (if ever) it will take to transition to such a civilization. After all, it will only be at the point where robots can create more robots, that a self-reinforcing loop emerges. At such a point, which could be thousands or millions of years into the future, much might also depend on politics (since as argued the economy itself disappears).
Nevertheless, the main (near-future) thesis is that AI has evolved far enough already (for proof, see the examples cited above of things where AI already excels at) such that a substantial amount of jobs could be automated in single-point use-cases, which will create demand for the Tesla robot, benefiting Tesla shareholders.
Financials
One could envision that in the future, every household could have a robot to offload everyday tasks to, just like pretty much every household has a PC or car. This means the potential market for such a device could (in the bullish case) could be just as large as the two examples just mentioned. The market for cars is roughly 100M in size, while 200-300M PCs are sold annually. Given the suggested price of $20k (although for consumer adoption the price might have to drop a bit), this means in the bullish case this new business could become nearly as large as Tesla’s existing EVs business (or even larger if Tesla reaches a higher marker share than in automotive).
Metaverse vs. utopiaverse
One of the drivers for Meta (META) creating the metaverse is the vision of how humans will work in the future. However, as Elon Musk has recognized, in the future there will be no work anymore (at least not in the strict sense). Although the metaverse may still be used for recreational purposes, Elon Musk has recognized that the potential for value creation – as argued, to the point where scarcity turns into abundance – in the real universe is far greater than in the metaverse.
Risks
While as argued Tesla has all the key ingredients in-house to build a successful robot business (in principle), as argued equally the usefulness and therefor economic value of the robot will ultimately depend on the software – not the physical hardware.
In that regard, the emergence of autonomous driving, or rather the lack of its emergence, as it has taken far longer than many, including companies, had expected, shows that the vision laid out here and by Tesla may not be as straightforward to actually execute on.
Hence, investors should recognize that Optimus is still in research.
Investor Takeaway
Elon Musk is absolutely right that the future of the economy is more and more automation and technology. Elon Musk has recognized that the most straightforward way to replace a human job is by creating a machine that simply looks like a human, infused with powerful silicon and software, both of which after well over 50 years of Moore’s Law belong the most abundant pieces of technologies that have been created.
The relatively recent rise of AI has made this possible only now. Whereas the past of technology was about the creation of tools to aid humans to be more productive (e.g., in farming), the future of technology will be to replace the middle man altogether – by middle humanoid robots.
After all, while economists most often tend to discuss things like inflation, and while politicians are most often concerned by how many jobs certain projects will create, investors should recognize that the real source of wealth over the last several hundred years has been solely due to technology, which (1) has increased the productivity of humans far beyond what any workforce, solely using ancient methods, would ever be able to accomplish, and (2) hence has caused an amount of deflation never seen before in the history of humanity. For example, if would take all of humanity (and then some more) in order to perform the calculations that the single device that you are using to read this is performing each second.
While a future where robots own and operate all infrastructure (including stocks) may still be far out, in the near-term there is a nonzero chance that Tesla’s new robot business could become a lucrative new business for Tesla, which could further improve its financials (and hence the stock) beyond its existing automotive, solar, storage and autonomous driving businesses.
Be the first to comment