Humanoid Robotics: The Economics of General-Purpose Labour

TL;DR

• Humanoid robots are moving toward becoming a potential multi-trillion-dollar industry, driven by robots with a human form factor that operate in the same environments and with the same tools as people.
• Advances in AI, lower hardware costs, and structural labour shortages in sectors such as logistics, manufacturing, and healthcare are making humanoids economically viable.
• The opportunity is a stack, not a single product: value will be distributed across intelligence and software, hardware and components, integration and manufacturing, and services.
• Competition is already underway between large corporations and heavily funded startups. However, adoption could still be constrained by safety concerns, regulatory constraints, and capital cycles, making humanoids a high-potential yet high-uncertainty theme for investors.

‍

In a very short time, humanoid robots have evolved from a science-fiction concept into a multi-billion-dollar global industry. Tesla is tying a large part of its future to Optimus, Chinese manufacturers are rolling robots into factories, and a handful of start-ups have raised multi-billion dollar rounds on the promise of developing the technology required for general-purpose robotic labour.

What makes this interesting is the underlying concept. If a robot with a human form factor can work in the same environment, with the same tools and complete the same tasks as a person, then a considerable slice of today’s work can technically be automated. This piece examines why humanoids are resurfacing, how they fit within the global economy, and the investment opportunities they present.

‍

What humanoids actually are?

At a high level, humanoid robots are machines built to mimic human appearance, movement and interaction. They are equipped with artificial intelligence, machine learning, computer vision and sensor technologies to perform human-like tasks. Crucially, the form factor enables them to operate in existing human-designed spaces with minimal modification and to draw on extensive data on human movement. That ability to integrate into the built environment is what distinguishes humanoids from traditional industrial robots or wheeled platforms, and why they are being positioned as a potential general-purpose form of robotic labour.

‍

Why are humanoids re-emerging?

On the technology side, AI has moved from narrow perception to broadly useful reasoning. Large language and multimodal models can now interpret images, text and sensor data and map that understanding to actions. ARK Invest frames humanoids as a means to translate foundation models into physical productivity, with a long-run annual revenue potential of approximately $24T, roughly split between industrial and household use.

Hardware has also caught up. Morgan Stanley models an average humanoid price falling from approximately $200K in 2024 to the low hundreds of thousands by the late 2020s, and potentially closer to $50K by 2050, as volumes grow and actuators, sensors, and compute power become cheaper. In China, where domestic suppliers already dominate the markets for motors, reducers, and batteries, several platforms are being announced at much lower price points. Layered on top of this is economic necessity. Many developed economies face ageing populations, tight labour markets and rising wages in logistics, manufacturing, healthcare support and defence.

‍

Where humanoids fit in the global economy?

The headline numbers for the market are eye-catching. Multiple sources indicate a global revenue opportunity exceeding $24T if it achieves full penetration, roughly split equally between household and manufacturing use cases. In the US alone, about 75% of occupations involve tasks that a humanoid could, in principle, perform, and about 40% of employees work in roles with meaningful “humanoidability”. If the robot can reliably deliver a small but persistent productivity edge at a predictable cost, the procurement decision begins to resemble a standard capital budgeting choice.

Labour data suggest that around three-quarters of occupations involve tasks that a humanoid could, in principle, perform, and roughly 40% of employees hold roles with meaningful “humanoidability”. In practice, early adoption is concentrated in work that is physical, repetitive and carried out in relatively controlled environments. Production, food preparation, transport and material moving, cleaning, installation and maintenance all fit that profile. Jobs built around human judgment or deep interpersonal interaction look much less exposed.

Geographically, North America and China are furthest along. The United States combines higher wages with large early adopters in ecommerce and logistics that are willing to trial new hardware in live operations. China offers scale, a dense component supply chain, and strong policy support, with domestic humanoid manufacturers already reporting meaningful orders for automotive and electronics plants.

‍

Creating a multi-layered investment opportunity

Humanoids are best understood as an ecosystem rather than a single product story. They create a stack of different risk and return profiles across the production chain.

• Intelligence and SoftwareThis is where perception, planning and control live. Margins here are high, and capabilities can be reused across many platforms, but access is primarily through large, publicly listed AI and semiconductor firms or a small group of highly valued private companies.
• Hardware and ComponentsThis layer covers actuators, gearboxes, motors, hands, batteries, sensors and frames. Precision component makers can enjoy strong pricing power, while commodity parts are fiercely competitive and heavily exposed to low-cost manufacturing regions.
• Integration and ManufacturingBig tech and auto groups are repurposing vehicle supply chains and factories to build humanoids. Pure play integrators are constructing dedicated plants and control stacks. Robotics incumbents are adapting existing platforms into humanoid form factors. This is likely to be the highest-revenue layer and the one in which price competition will be most visible.
• Deployment and ServicesThis section addresses integration work, training, maintenance, and Robot-as-a-Service pricing. Over time, a growing share of profit is likely to migrate into these recurring contracts rather than one-off hardware sales.

‍

Competition has already begun

On one side are large corporations that use humanoids to extend their existing businesses. Tesla is funding Optimus from its own balance sheet and has indicated that humanoids could become central to its long-term value. Hyundai plans to deploy commercial humanoids into its auto plants later this decade, targeting volume production rather than small pilots. Several other Chinese groups are also developing their own platforms.

On the other side are pure play humanoid startups. Some of these have already raised more than a billion dollars, with funds being used for factories, AI research and early deployments. Others have built dedicated humanoid robots in the United States and are deploying them into live warehouse environments in partnership with large logistics customers. Several are also partnering with frontier AI labs to strengthen the embodied AI layer. Apptronik, for example, positions Apollo for industrial deployments and highlights collaboration with Google DeepMind’s robotics work.

‍

What could slow adoption?

While the narrative is largely positive, there are good reasons to think humanoids may not scale as predicted. Hardware must transition from impressive prototypes to reliable, mass-produced machines that can operate safely and cost-effectively for years in harsh environments. Safety standards for bipedal robots operating alongside people are still evolving, so a high-profile incident could prompt stricter regulations, higher insurance costs, or bans in some settings.

Labour market and political responses will vary by country. Some governments may actively subsidise adoption in areas such as elder care and defence, while others proceed more cautiously or impose restrictions on the displacement of certain roles. Capital cycles matter too. The current funding environment has been highly supportive of ambitious hardware projects, meaning a broader downturn could pose obstacles for platforms that still require years of development capital.

‍

The Wealt Take

Wealt’s thesis in this piece is that humanoid robots are moving from exciting demos to deployable labour, and that the platforms best positioned to compound advantage are the ones that pair real-world deployment learning loops with credible scale pathways in manufacturing and embodied AI. Everything above supports that direction. Costs are trending down, labour shortages are pulling ROI conversations forward, and adoption is concentrating in controlled environments where safety and reliability can be proven.

Within that frame, Apptronik is a representative example of how the category is maturing. Apollo is positioned around practical industrial workflows, with an emphasis on operating safely alongside people, supported by a manufacturing-scale path via Jabil and a strengthened embodied AI stack through collaboration with Google DeepMind’s robotics work. Our view is that exposure to credible integrators with tangible scale paths offers a clean way to participate in the humanoid buildout as it moves from prototypes to fleets.

For Wealt investors, this is access to a late-stage private position in a leading emerging humanoid platform at a stage that would ordinarily require direct relationships with the company’s lead institutional investors. Wealt provides structured secondary access through terms and minimum sizes appropriate for qualified individual investors, offering targeted exposure to the integrator layer of humanoids as adoption and services revenue begin to compound.

At Wealt, we run independent research on companies shaping this transformation, giving our community access to opportunities historically reserved for institutions. If you would like to explore these insights or access upcoming allocations, you can apply to join the Wealt platform.

👉 Download the Wealt App on App Store or Google Play.

👉 Join the Wealt Club

‍

_{All investments carry risk, including the possible loss of capital, and past performance is not a guide to future returns. This opportunity is suitable only for sophisticated investors who understand and can bear those risks.}