Tesla has gutted the data annotation team working on Autopilot, laying off nearly 200 employees and shutting down the San Mateo, California office where they worked. The layoffs, first reported by Bloomberg, have been confirmed by sources who talked to TechCrunch on condition of anonymity.
The cuts come amid a broader reduction of jobs at Tesla. However, these layoffs targeted personnel once deemed critical to the company’s Autopilot advanced driver assistance system and more notably efforts by CEO Elon Musk to further develop automated driving functions through the $12,000 optional FSD system.
Until today, Tesla had hundreds of data annotation employees working on the Autopilot team in San Mateo and Buffalo, New York. The San Mateo office had a headcount of 276, and after laying off 195 staffers from all ranks — supervisors, labelers and data analysts — the team is left with 81 workers, who sources say will be relocated to another office.
Most of the workers were in moderately low-skilled, low-wage jobs, such as Autopilot data labeling, which involves determining if Tesla’s algorithm identified an object well or poorly, according to one source.
The source noted layoffs of this team were rumored to be on the table for months, and that the work would be offloaded to Buffalo.
Based on data from Glassdoor, jobs like data annotation specialists or data analysts at Tesla pay less in Buffalo than in San Mateo. It’s unclear if Tesla is shifting workers to the New York office to reduce costs or as a strategy to become eligible for New York State’s many job incentives, like the New York Youth Jobs program tax credit or the credit for employment of persons with disabilities.
That said, it likely won’t be a 1:1 replacement in Buffalo, where sources say Tesla will probably just inundate the existing team, “as is the Tesla way. Act now, deal with the consequences later.”
If Tesla is indeed not going to be hiring more data labelers and others to work on Autopilot — which the company has said is a vital ingredient for training deep neural networks that can help improve the full self-driving beta software — then what will happen to that technology? Perhaps Tesla will change course on its vision-only approach to autonomy and start introducing lidar and radar to its vehicles.
Layoffs or terminations?
While sources confirm the 195 Autopilot team members who were let go Tuesday were indeed laid off, they also say most of the “layoffs” that began in the end of May were, in reality, terminations based on performance.
Terminating employees based on performance allows any company the chance to avoid going through certain legal requirements like the Worker Adjustment and Retraining Notification (WARN) Act, which helps ensure advance notice in cases of qualified plant closings and mass layoffs.
Indeed, last week, two former Tesla employees filed a lawsuit against the automaker alleging the company did not provide the 60 days of advance notice required by federal law during its recent round of layoffs.
There is also a class action lawsuit being carried out against Tesla that sources say recruits more spurned workers everyday.