Elon Musk has promised that his company, Tesla, will launch “fully autonomous” cars in the next few years. Musk recently unveiled what he called a “Cyber Taxi” and said that Tesla plans to launch a robot taxi service by 2026, competing with other well-known brands currently operating in the space. However, “fully autonomous,” as Musk said, may be a bit of a misnomer. Recent reports indicate that the company is planning to hire a team of people to remotely troubleshoot its robotaxis.
In a recently spotted Tesla job listing, a vacancy has been announced for a remote teleoperation team for the company’s upcoming robotaxi fleet. “The Tesla AI Teleoperation Team is responsible for providing remote access to our Robotaxis and humanoid robots,” the ad states, emphasizing the additional need to assist Elon Musk’s new lineup of Tesla robots. “Our cars and robots operate autonomously in challenging environments. As we advance the artificial intelligence that drives them, we need the ability to remotely access and control them,” the announcement reads.
The job posting also notes that such a teleoperator center requires “creating a highly optimized, reliable, low-latency data stream over unreliable vehicles in the real world.” Teleoperators can be “transported” into the robotaxis using a “state-of-the-art virtual reality setup,” the announcement adds.
Tesla will not be the first company to use this method. In fact, it is an industry standard. It was previously reported that Cruise, a General Motors-owned robocab company, used remote human assistants to troubleshoot when its vehicles got into trouble (it seems that vehicles were in trouble every four to five miles). Google’s Waymo is also believed to be using the same practice, as is Zoox, a robotaxi firm owned by Amazon.
Overall, such methods seem to be part of a broader trend of companies positioning their products as “autonomous” or AI-powered, only to find that much of the work the product does is actually done by low-paid human contractors. Last year, it was reported that some of the key work on ChatGPT, OpenAI’s popular chatbot, was done by people who were paid as little as $15 an hour.
Silicon Valley wants us to believe that their autonomous products are some kind of self-directed magic, but the technology is clearly not ready yet. A quick glance behind the curtain consistently reveals a product base that, at the very least, is still heavily dependent on human labor.
