Late last year, attendees of an exclusive tech showcase in Burbank, California, bore witness to a true marvel of space-age innovation, a figure previously seen only in the far-off, fantastical worlds of science fiction – a walking, talking, autonomous humanoid robot. The robot (a crew of them, in fact) left guests awestruck as it danced, tended bar, played rock-paper-scissors and conversed with members of the crowd. The droid has a name – Optimus, a nod to the robo-alien protagonist of the “Transformers” franchise – and is the prototype for a general-purpose bipedal bot developed by automaker and energy company Tesla, host of the evening’s event, techno-optimistically dubbed “We, Robot.”
Tesla chief Elon Musk, not known for understating his company’s achievements, stirred up hype around Optimus, promising the “sentient” autonomous android could be “the biggest product ever of any kind,” eventually eclipsing the vehicles and batteries Tesla is known for. There’s just one problem – it’s not actually autonomous. At least not fully, and not yet. While Optimus dazzled audiences on the showroom floor in Burbank, elsewhere a team of human puppet masters was pulling the strings.
“Fauxtonomy” is a tongue-in-cheek term for tech products, from machines to chatbots, that are perceived as operating free of human control or intervention when in reality they rely at least partly on human operation. And it’s hardly an Optimus-only issue. Any salesperson will tell you that a degree of embellishment or exaggeration about one’s wares is simply how the game is played, especially when it comes to Silicon Valley. But the “We, Robot” gathering brought this somewhat unsavory phenomenon to the fore in the world of tech, with some observers rightfully feeling misled. Within a week of the attention-grabbing festivities, reputable sources including the Los Angeles Times reported that the Optimus prototypes had indeed been remotely operated, something Musk had conveniently failed to mention to audience members – including investors – when talking up the robots, which he eventually plans to sell to consumers for upward of $20,000.
The problem of “fauxtonomy” extends beyond the rarefied air of invite-only tech parties; in fact, it may already be rolling through the streets of your hometown. Auto companies like Waymo and Zoox, affiliated with tech giants Google and Amazon, respectively, already have robotaxis on the road everywhere from Seattle to Miami, with no one physically behind the wheel. But these futuristic cabs are not as “driverless” as they seem. While the vehicles may not have a human in the driver’s seat, they’re often under the close watch and frequent remote control of humans working behind the scenes.
As The New York Times notes, these autonomous taxi companies are hesitant to tell the whole truth that their vehicles are not wholly autonomous. Like Tesla, they know that a complex device appearing to operate entirely on its own generates far more buzz among consumers and investors than one openly receiving help from a team of humans. Cruise, a now-defunct autonomous taxi startup, reportedly had remote employees assisting its fleet at a ratio of 1.5 humans per vehicle, hardly an improvement on the single individual typically required to drive a car the old-fashioned way.
While in some cases the public might feel duped by human-controlled devices masquerading as autonomous technology, in other instances they may be grateful to learn that so-called autonomous systems involve human oversight – for now, at least. Take unmanned combat aerial vehicles (UCAVs), for example. The first use of UCAVs in acts of war dates back to a half-century ago; in recent years, however, AI-powered autonomous technology has enabled the rapid evolution of these drones to execute more and more functionality without immediate human influence.
Considering these UCAVs often have the capacity to take human lives with ease, it’s fair to say that we’re well on our way down a very slippery slope with regard to autonomous military drones. For the time being, the majority of UCAVs are piloted under human-in-the-loop or human-on-the-loop protocols – the first requiring a remote human operator to initiate the use of onboard weaponry, and the latter allowing a remote human operator to abort such an action. But the tech is trending toward more autonomy, not less, which means we may soon have to confront the grim reality of self-governing automatic killing machines.
Since 2018, the UN’s Office for Disarmament Affairs has considered lethal autonomous weapons systems (LAWS, a category that would include fully autonomous combat drones) to be “politically unacceptable and morally repugnant” and pushed for their prohibition. While not thoroughly and consistently defined, LAWS are essentially human-out-of-the-loop systems, whereby a target could be identified and eliminated without any direct human input. But there are those who feel prohibition is futile, and that late adopters of such systems could be at a strategic disadvantage. This past April, UK politician and former Royal Marines Colonel Alistair Carns released a statement urging his government to accept the coming dawn of fully autonomous combat drones, a move that would require amending the country’s current legislation restricting LAWS.
Whether in reference to android assistants, robotaxis, or military drones, the term “autonomy” has been applied liberally to various sectors of developing technology, without drawing clear lines between systems and devices that are primarily under remote human control, largely automated with occasional human intervention, or fully self-sufficient. Even those technologies that independently use advanced AI for complex decision-making and the completion of tasks inevitably have been influenced by humans in the programming and training phases. Plus, if you ask philosophers to weigh in on the question of free will, even humans themselves arguably can’t be considered fully autonomous. So with this much ambiguity around the parameters of autonomy, can we really expect a straight answer on the subject with regard to our technologies?
Speaking of not getting a straight answer, let’s check back in with Elon Musk. On Tesla’s Q4 earnings call this past January, Musk made a bold proclamation about his company’s deployment of autonomous tech in its auto products. He announced that Tesla vehicles, which thus far have featured “fauxtonomous” full self-driving (FSD) mode only under the active supervision of the driver, would launch unsupervised FSD this year. Tesla’s X account even posted, “The future is autonomous & it starts in Austin, this June.” So here we are in June – has the autonomous future started?
Turns out Tesla is simply planning to roll out a fleet of driverless robotaxis, which, if you ask the likes of Waymo, Zoox, et al., is a future that started years ago. Plus, if those self-driving competitors are any indicator, Tesla’s robotaxis, just like Optimus back in Burbank, will be operating under the watchful eye and intervening hand of a human chaperone. With AI-enhanced technology looming ever larger on the horizon, poised to take over human tasks from social interactions to hard labor and everything in between, professional buzz-builders like Musk tout the vision of a near future in which human beings fully hand over the reins to our inorganic counterparts, which, you guessed it, will be made and sold by the buzz-builders themselves. Until further notice, however, that vision is under supervision.
