Are humanoid robots really the future?

As companies like Figure and Tesla race to build general purpose humanoid robots, it's still not clear whether they're onto something, or merely caught up in a sci-fi fantasy fever dream.

Figure 02 (left) and Optimus (right)

If you’d asked me a few years back whether we’ll be living side by side with humanoid robots in the near future, I’d of laughed. But a growing number of companies are betting on a future where we work, live, and play, with very human-like robot colleagues and companions.

The trouble is, they may still be living in a science fiction-fueled fantasy.

This past week one of the fastest-growing humanoid robot companies — Figure — released details of its next generation robot, Figure 02:

Advertised as the “world’s first commercially-viable autonomous humanoid robot”, Figure 02 is designed to be a human-replacement in workplaces, lending “support across manufacturing, logistics, warehousing, and retail.”

Working in collaboration with OpenAI, Figure 02 is designed to be a general purpose humanoid robot that can communicate, reason, and learn on the job (in the launch video above it’s shown being tested on a BMW production line).

A number of other companies are also working on human-like robots — perhaps the best known one being Boston Dynamics with it’s it’s flagship humanoid robot Atlas and a range of robots designed to carry out specific tasks.

But out of all the competition, the product that comes closest to Figure 02 in design, function, and aspirations, is Tesla’s Optimus.

Tesla released the second generation of the Optimus robot late last year, and there are rumored plans for a completely redesigned third generation later this year, or early next.

Both Figure’s 02 and Tesla’s Optimus are interesting in that they’re being developed on the back of a very specific vision of the future. This is one where robots replace manual human labor (and fill a perceived shortfall in the human workforce); radically increase manufacturing output; transform lives by taking on mundane, difficult, and essential tasks; and ultimately take humanity to other planets.

As Brett Adcock, founder and CEO of Figure, puts it:

Today, we are seeing unprecedented labor shortages. There are over 10 million unsafe or undesirable jobs in the U.S. alone, and an aging population will only make it increasingly difficult for companies to scale their workforces. As a result, the labor supply growth is set to flatline this century. If we want continued growth, we need more productivity — and this means more automation.

…

We believe humanoids will revolutionize a variety of industries, from corporate labor roles (3+ billion humans), to assisting individuals in the home (2+ billion), to caring for the elderly (~1 billion), and to building new worlds on other planets. However, our first applications will be in industries such as manufacturing, shipping and logistics, warehousing, and retail, where labor shortages are the most severe. In early development, the tasks humanoids complete will be structured and repetitive, but over time, and with advancements in robot learning and software, humanoids will expand in capability and be able to tackle more complex job functions.

Elon Musk has similar aspirations for Tesla’s Optimus, along with an (admittedly optimistic) projection that there could be a market for a billion plus humanoid robots per year in the future!

It’s a figure that Tesla’s future as a company is depending on — at least in part — and is one that represents a massive bet on a future that is far from certain.

Both Figure and Tesla believe that the secret sauce of a successful humanoid robot future is to build machines that understand and interact with a world made for humans and inhabited by them.

For this, they argue, their robots need to be able to maneuver like humans across an incredibly diverse range of environments; they need to be able to perceive their surroundings as a human would — hence the focus on optical sensors that mimic human sight; they need the manual dexterity of a human (and this is where much of the new tech in Tesla’s Gen 3 Optimus is being focused); they need to be able to learn from experience, like humans; and they need to be able to communicate with humans in the same way that other humans do.

These design parameters are what are driving the development of Optimus and Figure and that will lead — if successful — to mechanical human replacements.

If you were brought up on science fiction like Asimov’s robot novels and short stories, or a whole raft of robot-centric science fiction movies, this is a vision that I suspect feels like manifest sci-fi destiny — something my colleague Rizwan Virk calls the “sci-fi feedback loop.”

But I suspect that it will take more than visions of sci-fi futures past to bring about a world where humanoid robots are ubiquitous.

This is not necessarily because of technical challenges — which are fiendishly difficult, but not intractable — but because it takes a lot more than cool tech alone to make a product successful.

When it comes to general purpose humanoid robots, there are certainly potential economic barriers. These include balancing efficiency gains versus costs over time, especially as it’s yet to be shown that human-replacement robots can increase efficiency and decrease costs compared to a combination of human labor and task-specific machines.

But the biggest barriers are likely to be societal rather than technological. And, perhaps, to a lesser but nevertheless important degree, environmental.

And the biggest question here is: Will people tolerate mechanical human replacements?

Both Figure and Tesla are initially looking to introduce their robots to workplaces as a general purpose human replacement on production lines and elsewhere. Superficially this could lead to efficiency and productivity gains as companies reduce their need to pay people or provide leave, health, and retirement benefits.

But if the displacement is substantial — or it looked as if it might be — there will likely be a very large backlash as peoples’ jobs are threatened.

Of course, modern history is replete with examples of automation taking jobs — followed by protests, but ultimate acceptance. But given the scale at which proponents of humanoid robots are thinking, my sense is that the prospect of tens to hundreds of millions of robots taking human jobs will be disruptive to the point of being challenging to implement.

And it’s the general purpose nature of humanoid robots that will fuel this as people realize that, unlike previous waves of automation, these robots are more able than most workers to learn new skills and adapt to emerging needs.

In other words, I’d not be surprised to see a Luddite-like backlash against humanoid robots in the workplace that, unlike the original Luddite movement, actually succeeds.

If this occurs, such a backlash will be driven by job losses and job displacement. But I suspect that safety concerns will also feature prominently in it.

At this point, we simply don’t know how safe humanoid robots will be when operating in close proximity to or with with people. I’m pretty sure companies like Figure and Tesla are working hard on this. But the reality is that — just as we’re seeing with self-driving cars — it’s going to take millions of hours of simulation and training to be sure that humanoid robots are acceptably safe.

And given that the modes of interaction will have many more degrees of freedom compared to autonomous vehicles, my guess is that the bar for demonstrating acceptable safety is going to be a high one.

This doesn’t mean that these robots won’t be safe. But it’s the edge cases — the unexpected — that will ultimately set the bar. Especially as safety is deeply grounded in a perception of what is considered to be acceptably safe — which is highly subjective.

This doesn’t bode well for the future of such robots in the workplace. But things get even more challenging when considering a future where everyone has their own humanoid robot.

This is certainly part of Tesla’s vision, and I suspect it underpins plans to sell billions of units a year in the future. But if the safety bar is going to be high in the workplace, it’s going to be stratospheric in the home.

And it’s a bar that will be underpinned by perceptions and assumptions of safety — and perceived threats to what is considered sacrosanct — rather than hard evidence.

For people to invite a humanoid robot into their lives, they will need to feel safe and comfortable with it. They will need to know for certain that it will never harm a child, or a pet, or a visitor. And they will need to be sure that it will never — under any circumstances — make them feel insecure.

To illustrate this, imagine the scenario where you’re can’t sleep and head for the kitchen in the middle of the night for a drink —just to be startled by a glimpse of an unexpected human-like figure lurking in the shadows. Or you wake up to see your humanoid robot seemingly watching you from the end of the bed.

Neither of these scenarios presents a physical risk. But the chances of being spooked to the extent that you cannot feel safe and secure in your own home until the device has gone, is not negligible. And as this is a response that would be driven by biology and instinct, not rational thought, it’s not one that’s easy to overcome.

These are just two examples of many that underline the challenging psychology of accepting mechanical humans into our lives.

Of course, there will be early adopters who will revel in the idea of a robotic companion or colleague. And there will no doubt be solid arguments for purchasing a robot home-help (although I also think we’ll discover that part of being human is having the autonomy to do the “boring” tasks that life is full of, rather than being relieved of these).

But it’s hard to imagine a future society where our biological tendencies to react instinctively and negatively to something that is uncannily human invading our space, does not severely limit the use of humanoid robots.

I may be wrong. But either way, I’m pretty sure that the future success of humanoid robots will depend primarily on the deeply complex challenge of ensuring human acceptance, rather than the merely complicated challenge of making them in the first place.

Of course, a smart pathway forward for companies like Figure and Tesla would be to build robust teams that focus on overcoming the societal challenges they face.

I’m just not sure they realize this yet.