How ARK Is Thinking About Humanoid Robotics

Introduction

As Elon Musk said during Tesla’s first-quarter earnings call, “If you've got a sentient humanoid robot that is able to navigate reality and do tasks at request, there is no meaningful limit to the size of the economy.”¹ Musk was sharing a vision that we believe is turbocharging the robotics industry, creating new robotics companies and increasing venture investments focused on its promise.  

While Musk is envisioning humanoid robots that exceed human capability, even before that moment in time, if humanoid robots are able to operate at scale, they could generate ~$24 trillion in revenues, split roughly equally between household and manufacturing robotics, according to ARK’s research, as shown below.

*Note: the cells highlighted in green represent what ARK believes to be a reasonable or likely outcome. Sources: ARK Investment Management LLC, 2024. This ARK analysis is based on a range of underlying data from external sources, which may be provided upon request. Forecasts are inherently limited and cannot be relied upon. For informational purposes only and should not be considered investment advice or a recommendation to buy, sell, or hold any particular security.

In this article, we focus on the robotics opportunity in US manufacturing. According to our research, the US manufacturing sector employs nearly 12 million people, who work ~23 billion hours per year, for ~$785 billion in pay, to produce output worth ~$2.4 trillion, as shown below. In the unlikely event that it were to substitute robots for all human workers and put them to work for 16 hours per day, the manufacturing sector would need only ~5.9 million robots—half the number of human workers employed today—to deliver the same level of manufacturing output, as shown below.

US Manufacturing

*This figure assumes a ~50% take rate for the provider of the humanoid robot ($785 * 50% = $392.5). Source: ARK Investment Management LLC, 2024, based on data from the Bureau of Economic Analysis, U.S. Department of Commerce 2024 and United States Census Bureau 2023.²  For informational purposes only and should not be considered investment advice or a recommendation to buy, sell, or hold any particular security.

In a more likely scenario, humanoid robots will debut at higher price points and be much less capable than their human counterparts. According to our research, humanoid robots will become more economically viable than human employees at tipping points in their net present value (NPV) that balance the cost of humanoid robots against the productivity gains they will enable, as shown below. At a cost of $16,000, for example, a humanoid robot would have to deliver little more than a 5% gain in productivity relative to its human counterpart to become economically viable. For context, Musk has suggested that, “Complexity per unit mass is much higher with humanoid robots, but still I think it ends up costing less than half of a car.”³

Source: ARK Investment Management LLC, 2024. This ARK analysis is based on a range of external data source, which may be provided upon request. For informational purposes only and should not be considered investment advice or a recommendation to buy, sell, or hold any particular security.

As we put the tipping point model into real-world context, two variables in manufacturing are important to consider: company size and “labor share”—the share of revenue going to compensate labor.

Company Size

Unlike in small firms—where employees are likely to “wear more than one hat”—large manufacturing firms are organized by specialized and automated tasks. As a result, over time their productivity has been more than twice as high as the smallest firms.⁴ Specialization and automation give large firms the wherewithal to scale significantly and, in turn, lower labor costs as a share of revenue. Consequently, and somewhat counter-intuitively, large companies typically pay higher wages than small firms, because automating specific tasks typically boosts the productivity in large firms more than in small firms, as illustrated in the two charts below. 

Source: ARK Investment Management LLC, 2024. This ARK analysis is based on a range of external data source, which may be provided upon request. For informational purposes only and should not be considered investment advice or a recommendation to buy, sell, or hold any particular security.

Source: ARK Investment Management LLC, 2024, based on data from United States Census Bureau 2023.⁵ For informational purposes only and should not be considered investment advice or a recommendation to buy, sell, or hold any particular security.

Because generalized automation solutions—those for multiple tasks—have not evolved as quickly as automation solutions for specific tasks, small firms typically have a disproportionate number of automatable-but-not-yet-automated tasks that would benefit from generalizable solutions like humanoid robots.

Labor Share

Compensation as a share of revenue varies dramatically by subindustries in the manufacturing sector. In 2021, it ranged from 3% in tobacco manufacturing and 6% in grain and oilseed milling to 37% in machine shops (including turned products, and screws, nuts, and bolts) to ~40% in apparel knitting mills, as shown below.

Source: ARK Investment Management LLC, 2024, based on data from U.S. Bureau of Labor Statistics as of July 2024.⁶ For informational purposes only and should not be considered investment advice or a recommendation to buy, sell, or hold any particular security.

Thanks to their more specialized employment positions, large manufacturers typically benefit more than small companies from single task automation solutions and enjoy lower labor costs relative to revenues, as shown below. With less job specialization, small manufacturers are burdened with higher labor costs and, therefore, are likely to benefit disproportionately from more generalizable humanoid robotic solutions.

Source: ARK Investment Management LLC, 2024, based on data from  IZA Institute of Labor Economics 2019.⁷ For informational purposes only and should not be considered investment advice or a recommendation to buy, sell, or hold any particular security.

According to our research, 40% of US manufacturing employees work in small firms. In the illustration below, small firms employ fewer than 500 people.

Source: ARK Investment Management LLC, 2024, based on data from United States Census Bureau 2023.⁸ For informational purposes only and should not be considered investment advice or a recommendation to buy, sell, or hold any particular security.

Potential Tailwinds For Humanoid Robotics

In the aftermath of COVID-related supply chain shocks, onshoring and labor shortages could provide meaningful tailwinds for humanoid robots, lowering managements’ sensitivity to their price and, thereby, accelerating the transformation of manufacturing. As a result, our research suggests that the market value of generalizable robots could scale into the tens of trillions of dollars. 

The biggest question now is, how quickly will AI software enable humanlike performance across the subsectors of manufacturing, and beyond?