+
Generic selectors
Exact matches only
Search in title
Search in content
Post Type Selectors

2024’s Humanoid Hiring Spree and the Real Value of Work

Two years ago, ChatGPT sparked a global discussion about A.I. displacing creative professionals and other white-collar workers. Now, a developing class of autonomous humanoid robots is sparking similar concerns about manual labor in factories, warehouses, and other blue collar workplaces. 

2024 is shaping up to be the year of the robot. 

Nvidia CEO Jensen Huang said at the company’s annual computing hardware technology conference last month that “the ChatGPT moment for robots might be around the corner.”

Advances in A.I. software, computing hardware, and electric motors are fuelling a growing wave of startups looking to put humanoid robots in car factories, fulfillment centers, and even nursing homes. 

At the same conference, Nvidia unveiled “Project GR00T,” a software platform that enables humanoid robots to replicate human behavior by processing human language and video input. The company also announced a powerful new system-on-a-chip that can run GR00T and other models locally inside a robot. These breakthroughs mean that robots don’t need to be hardcoded in advance, and can adapt much more easily to changing environments. 

Robot “bodies” are also built for greater independence than ever before. Hyundai-backed robot lab Boston Dynamics recently retired their Atlas robot; the dancing, acrobatic, and often clumsy hydraulic-powered humanoid that dazzled and amused the lab’s social media audience. A lab in China just showed us the future: a robot built with powerful electric motors that can do backflips just like Atlas and that recently broke the humanoid speed record at a pace of 7.4 MPH. Battery packs for several humanoid robot models can also last upwards of four hours. 

These developments are driving a rapidly growing industry of robotics labs as well as difficult conversations about the future of work. How do we balance human labor protections with innovation, how do we structure taxes to reflect a changing workforce, and how do we design our economy so the benefits of robot labor extend beyond the robots’ owners and manfucturers? 

Image: Nvidia

Car Manufacturers Test-Drive Humanoid Robots

This year has already seen a flood of investments and new partnerships formed between robotics labs and industries with large numbers of manual laborers, such as car manufacturing and logistics.

In January, BMW announced a deal with a two-year-old robotics lab to test humanoid robots at their manufacturing facility in South Carolina. The same lab, Figure, recently raised $675 million from investors, including Jeff Bezos, Microsoft, Nvidia, and Intel. OpenAI is also partnering up to develop specialized A.I. models for the “Figure 01” humanoid. 

Although Figure doesn’t have a viable product yet, the company released a demo last month showing the upper half of Figure 01 understanding the implicit context of its environment through computer vision, and executing a series of steps to complete a broader task. In the demo, the robot offers an apple when asked for something to eat, identifies garbage to clean, and recognizes that the dishes need to be stored in the drying rack based on the way they’re placed. Robots that can understand instructions as easily as humans mean that companies don’t need to invest in technology experts to operate them, eliminating a massive barrier to robot adoption. 

Figure CEO Brett Adcock told the Associated Press last year, “If we can just get humanoids to do work that humans are not wanting to do because there’s a shortfall of humans, we can sell millions of humanoids, billions maybe.”

Figure’s competitors are making the same bet.

“Apollo” humanoids built by Texas-based Apptroniks are already working on the manufacturing floor at a Mercedes plant in Hungary. Another European company that manufactures Mercedes, Jaguar, and BMW vehicles recently announced a partnership this month to pilot Sanctuary AI’s “Phoenix” humanoid. 

Tesla is also making progress on its “Optimus” humanoid, sharing videos of the generation 2 model walking more steadily and displaying a wider range of movement. 

All signs point to us being on the cusp of a paradigm shift in the market for human labor, yet we’re still early in the process of deciding how we adapt our laws and culture to ensure that we all reap the benefits of automated labor.

Image: Helen Todd

The Promise and Challenges of Protecting Human Labor with A Robot Tax  

During his 2020 campaign for U.S. President, entrepreneur Andrew Yang generated early buzz for his platform with a plan to charge a tax on robot productivity and use the revenue to give every American a universal basic income (UBI) of $1,000 per month. 

Yang’s not the only one to endorse a robot tax; Senator Bernie Sanders, Bill Gates, Elon Musk, Mark Cuban, and others have voiced support for addressing the socioeconomic risks of labor automation with tax policy. 

Sanders is among the few who see a robot tax as a way to protect human labor by stopping or slowing the transition to automated labor. Gates and others see a robot tax as a way to soften the negative consequences of an automation revolution that’s sure to come. Protecting human labor is only one of the potential benefits of a robot tax, though. 

British law professor Ryan Abbot fleshed out the central argument for a robot tax during a 2019 debate hosted by MIT Technology Review. He argued that not taxing robots is actually equivalent to us subsidizing our replacement. Currently, the government funds social services like Medicare with revenue from payroll taxes. If less humans work, then funding those programs becomes more difficult. Robots, on the other hand, actually reduce their owners’ tax burden because their value (and, therefore, the overall value of the company) decreases over time. As taxes are usually one of a company’s biggest expenses, reducing that liability is high on any executive’s to-do list. 

Social services are already struggling to meet the demands of an aging population; a tidal wave of unemployment would trigger a crisis where demand for public benefits suddenly skyrockets, and at the same time, the funding source for those benefits is redirected to private corporate shareholders. 

At the least, Abbot suggests overhauling the corporate tax code to reduce the tax incentive of robot automation versus human labor. With a more balanced tax policy, it’s harder for companies to enrich themselves by replacing humans at the expense of social welfare. 

Despite the clear benefits of a robot tax in leveling the playing field for human workers, recent research shows it would be challenging to implement. 

Rosanna Merola, a macroeconomist and researcher at the International Labour Organization, published a paper last year concluding that, while a robot tax is “philosophically appealing,” implementing one incorrectly could be disastrous. Among the greatest challenges is the deceptively simple question of how we define a robot because when we talk about robots, we’re often actually talking about automation in general. Does an LLM or a digital avatar count as a robot? These decisions could have unintended consequences, like punishing companies that automate tasks without displacing human workers. 

Last year, a team of MIT economists published the findings of a study into how high a theoretical robot tax would need to be in order to replace the revenue lost from job automation. Perhaps to Bernie Sanders’ disappointment, the study found that a modest tax between 1 and 3.7 percent of a robot’s value is sufficient to replace the revenue losses. 

The other piece of the puzzle is what we do with the revenue from a robot tax. Depending on how aggressive the tax is, it could fund retraining for displaced workers or similar initiatives, it could go towards increasing unemployment benefits for people replaced by robots, or it could go towards a universal basic income as the Yang ‘20 campaign suggested. 

Opponents of UBI balk at the idea of paying people regardless of whether they work, arguing that UBI would cultivate a society of laziness and entitlement. The real-world results of UBI tell a different story, though. After all, it’s a public policy concept that’s been proposed in various shapes and sizes for centuries. 

Thanks to large-scale studies in Europe and Africa, we have the data to see what happens when people are given unconditional cash transfers. 

The nongovernmental organization GiveDirectly is in the middle of running the world’s longest and largest UBI programs over a 12-year period with 5,000 participants in Kenya. With a UBI of less than a dollar a day, over the past five years participants in Kenya consistently left low-wage jobs and started businesses that succeeded because their neighbors could afford to shop. Quality of life indicators such as nutrition and crime rates improved as well. Other studies by GiveDirectly have shown that cash transfers lead to better quality-of-life outcomes for recipients than alternative aid methods like nutrition vouchers or employment training. 

In the Netherlands, a research team provided social welfare beneficiaries with a basic income for two years. While some participants reduced their working hours, others used the income cushion to find permanent work and escape a cycle of low-wage gig work. According to the study’s author, both outcomes could be considered good in a world where there are not enough jobs for everyone. 

“Why try to push everyone into paid work?” Groot said in an interview with The Guardian. “Why not give people who have good alternatives the opportunity to reduce work, or not work at all, and cash in the basic income?”

While implementing a robot tax would be challenging and funding a UBI program would require overcoming even bigger cultural divides, the research and data show that both are possible and reasonable approaches to addressing the trend of robots replacing human workers before it becomes a crisis. 

For many people, though, their job is a part of their identity, and working is what drives them forward. They don’t want to lose their job to a robot, even if it means they get to enjoy a life of leisure. Cultures and religions are built on the merits of work ethic and in many cultures, the chronically unemployed are outcasts. It’s worth noting that not all cultures find meaning and purpose in their lives tied to work.

The laborers of love shouldn’t worry too much though, as the robot worker revolution is far from certain. 

Image: Apple

The Coworker That Never Goes Home

There’s evidence that robot coworkers will complement and expand human labor rather than replace it. 

A trio of unrelated studies published over the past five years in Canada, France, and Spain each measured growing employment at firms that adopted robot automations. Meanwhile, firms in the same industries that did not adopt robots saw employment slow or decrease. 

One theory is that the free market allocates more work to the more efficient firm, allowing those firms to hire more people. In that case, the best form of job security is actually working at a firm that’s friendly to automation. 

Research shows that you might even like that job better than your counterparts at firms that don’t automate. Three years ago, a study in Japan showed that robotic assistance reduced staff turnover at nursing homes by assuming physically demanding tasks and freeing up staff to focus on residents. 

In a 2017 interview supporting the robot tax, Bill Gates outlined a similar vision for our working relationship with robots. 

“What the world wants is to take this opportunity to make all the goods and services we have today and free up labor – let us do a better job of reaching out to the elderly, having smaller class sizes, helping kids with special needs,” he said. “All of those are things where human empathy and understanding are still very unique, and we still deal with an immense shortage of people to help out there.”

While there are valid fears about the consequences of robot labor for society, there’s also good reason to be excited about the important work that robots will enable us to do with – and for – our fellow humans. 

Nonetheless, it’s important that we protect the opportunity to work for those who are able and willing by eliminating the tax incentives for companies to eliminate human jobs. In doing so, we can ensure that, regardless of the degree to which robots replace us, our society is prepared, at the very least, to support those who get replaced. At best, we earn more freedom to choose our work and the conditions under which we do it, knowing that we can afford to.