What did Musk say about China and autonomous driving?
On Sunday, April 28, Tesla CEO Elon Musk visited Beijing.
This was an unannounced visit, and on the business front, speculation suggests he came to promote the launch of Full Self-Driving (FSD).
China is Tesla's second-largest market after the U.S., but sales have been declining this year. Tesla introduced FSD four years ago as the most autonomous version of its Autopilot software, though it has yet to launch in China, where only lower-level autonomous driving features are available.
$Tesla(TSLA.US)
Earlier this month, Musk hinted that FSD might be made available to Chinese customers "soon." Tesla recently halved the price of FSD in the U.S. to attract more users.
To enhance FSD capabilities, Tesla also needs data from China to train its technology. Tesla's vehicle data in China is stored in a domestic data center built in 2021.
Since 2024, Tesla's stock price has dropped by nearly one-third, while local Chinese competitors have grown rapidly. Tesla has cut thousands of jobs (over 10% of its workforce) to adapt to the new landscape.
In Q1, Tesla's revenue fell 9% year-over-year to $21.3 billion due to lower vehicle prices and deliveries. Profits hit their lowest level since 2021, with operating margins dropping from 11.4% to 5.5% compared to the same period last year.
Five days before his Beijing visit, during Tesla's earnings call, Musk discussed his views on Chinese competitors, the timeline for new vehicle launches, his belief in autonomous driving, and the vision for robotaxis. Musk compared Tesla's planned fleet operations to Airbnb. Tesla is set to unveil its robotaxi on August 8.
"If anyone doesn’t believe Tesla will solve autonomous driving, they shouldn’t be investors in the company."
"In the future, non-autonomous gasoline cars will be like riding horses and using flip phones."
"Even if I were abducted by aliens tomorrow, Tesla would still solve autonomous driving—maybe slower, but it would get done."
"If a company makes 5% organizational mistakes annually, it accumulates into 25% or 30% inefficiency."
In April, Musk postponed his trip to India, where he was scheduled to meet Prime Minister Narendra Modi. Reports suggest Tesla plans to invest around $3 billion in building a factory in India to produce lower-priced cars and begin imports later this year.
Excerpts from Musk’s Q&A with analysts on April 23:
As we’ve all seen, global EV adoption faces pressure, with many automakers scaling back EV production in favor of plug-in hybrids. We believe this is the wrong strategy—EVs will ultimately dominate.
Despite challenges, Tesla’s team excelled in energy storage deployments, with Megapack hitting record highs in Q1, driving the energy business to its most profitable quarter ever.
This momentum will likely continue, growing faster than the auto business in coming quarters and years.
In Q1, we expanded AI training capacity, doubling compute resources.
There’s been much discussion about upcoming models. We’ve updated our lineup to accelerate launches, potentially starting production in early 2025 or even late 2024.
New, more affordable models will use next-gen and current platforms, producible on existing lines without new factories.
At full capacity, we expect output exceeding 3 million vehicles annually.
FSD V12 is pure AI autonomy. If you haven’t tried it, we highly recommend it—it’s transformative and improving rapidly. It’s now enabled on ~1.8 million cars in North America with Hardware 3.
About 50% of users are actively using it, with adoption growing weekly. FSD V12 has driven over 300 billion miles.
Vision-based neural networks are clearly the right solution, mirroring human driving. We’ve lowered the subscription to $99/month to boost accessibility.
We’ll unveil the Robotaxi (Cybercab) on August 8.
On AI compute: We’ve deployed 35,000 H100s and expect ~85,000 by year-end for training alone.
FSD V12 proves human-level reliability is imminent. The future is EVs and autonomy.
As I’ve said, non-autonomous gas cars will soon be like horses and flip phones.
(On Optimus humanoid robot mass production timeline?)
We’ve demonstrated simple factory tasks in labs. Optimus may enter limited production by late 2025. It could eventually surpass Tesla’s total value.
(FSD safety and progress?)
Think of it like elevators—you don’t need to know how they work. Our fleet model blends Airbnb and Uber: owners can share cars selectively.
At scale (millions then tens of millions of vehicles), continuous feedback will create a moat like Google’s search dominance.
With Hardware 5 expected by late 2025, a future 100-million-vehicle fleet could offer 100+ GW of distributed compute capacity.
($25K car timeline?)
More details coming August 8. But Tesla’s true focus is solving autonomy—the ultimate value unlock.
By decade’s end, we expect tens of millions of Tesla vehicles on the road.
(Plans to reduce Tesla involvement?)
Tesla consumes most of my workweek. I’m ensuring its prosperity now and ahead.
(Competition in China?)
We’re outperforming rivals. As Cathy Wood noted, Tesla is an AI/robotics company—viewing it as just an automaker is fundamentally wrong.
Autonomy is inevitable—with or without me.
(On layoffs?)
Post-2019 growth required restructuring. Like organisms, companies must evolve structures to scale—or fail.
(FSD in China?)
We plan to launch FSD in China pending regulatory approval. Location-specific training will optimize performance, but it works nearly everywhere.
The copyright of this article belongs to the original author/organization.
The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.
