Connect with us

Tesla

Tesla FSD improvements in China continues with Baidu’s help

Tesla is working with Baidu to improve FSD in China, as strict data laws prevent Tesla from training its AI locally.

Published

on

Credit: Tesla China/Weibo

Baidu is helping Tesla Full Self-Driving (FSD) improve in China.

Engineers from Baidu, a leading AI company in China, are helping Tesla with its FSD Version 13 software.

According to Reuters’ sources, Baidu recently dispatched engineers from its mapping team to Tesla’s Beijing office. The engineers are working on integrating Baidu’s navigation map information with Tesla’s FSD V13 software.

Tesla’s way of improving FSD V13 in China differs from the methods used in the United States. In the U.S., Tesla trains FSD to improve navigation techniques and learn about roads. In contrast, Tesla cannot train FSD in China due to the country’s data laws.

“Then China, which is a gigantic market, we do have some challenges because they currently don’t allow us to transfer training videos outside of China. And then the US government won’t let us do training in China. So, we’re in a bit of a there. It’s like a quandary,” noted Elon Musk during the last earnings call.

“So, we were solving then is by literally looking at videos of streets in China that are available on the Internet to understand and then feeding that into our training so that publicly available video of street signs and traffic rules in China can be used for training and then also putting it in a very accurate simulator. And so, it will train using SIM for bus lanes in China,” he added.

Advertisement
-
-

Tesla has rolled out FSD (Supervised) in Mexico and started the process to release it in parts of Europe. Despite a few regulatory issues, Musk is optimistic that FSD (Unsupervised) will be available in more countries this year.

“But I think we’ll have unsupervised FSD in almost every market this year, limited simply by regulatory issues, not technical capability. And then unsupervised FSD in the U.S. this year, in many cities but nationwide next year. And hopefully, we have unsupervised FSD in most countries by the end of next year,” he noted during the Q4 and Full Year 2024 earnings call.

Maria--aka "M"-- is an experienced writer and book editor. She's written about several topics including health, tech, and politics. As a book editor, she's worked with authors who write Sci-Fi, Romance, and Dark Fantasy. M loves hearing from TESLARATI readers. If you have any tips or article ideas, contact her at maria@teslarati.com or via X, @Writer_01001101.

Advertisement -
Comments

News

Tesla’s AI lead doubles down on FSD’s speed strategy, and owners are confused

Published

on

Credit: Tesla

Tesla’s AI lead Ashok Elluswamy doubled down on the company’s strategy regarding Full Self-Driving’s speed settings, and owners are definitely confused.

Earlier versions of Full Self-Driving allowed owners to set a max speed that the vehicle could travel while operating under the semi-autonomous driver assistance platform. This allowed more customization for the driver, giving them the ability to experience FSD’s robust performance with their own personal preferences.

Speed is massively important for obvious reasons — it’s not only a question of keeping the vehicle occupants comfortable by traveling at a safe speed, but it’s also something that could contribute to a ticket or infraction from law enforcement.

With the release of FSD v14 last year, Tesla removed the ability to set a max speed and instead opted for five Speed Profiles, ranging from “Sloth,” the most conservative, to “Mad Max,” the most aggressive and spirited. These profiles not only control speed, but also how frequently the vehicle will execute passes, perform lane changes, and other contributing factors.

The removal of the Max Speed setting was a major complaint amongst the Tesla community because it left owners scrambling for a way to experience suitable behaviors while traveling at an appropriate speed. Most felt the driving profiles would be a good indicator of the behaviors, while speed would still be left up to the discretion of the driver.

Instead, Tesla’s Speed Profiles determine both, and the constant tinkering of how they behave has been a major bottleneck and point of confusion for both owners and the company. From update to update, the Speed Profiles will change, sometimes more drastically than others. Some owners have complained that the “Standard” profile is too fast, while others have experienced “Mad Max” traveling below the speed limit:

These things change with each update, but the big complaint is that owners are on the hook for any tickets that come from FSD’s infractions; that’s the caveat of the suite being named FSD (Supervised). It ultimately means the driver is responsible, and the automaker has no liability when it comes to speeding tickets or general traffic infractions.

It is the driver’s responsibility to take over or adjust based on this.

Elluswamy essentially confirmed that there are no plans to bring back Max Speed control, because it is what he referred to as “an anti pattern.” He then echoed something that CEO Elon Musk has started to really push with FSD, and that’s the idea that Tesla is really honing in on the preferences of the driver.

Owners were confused by Tesla’s decision, stating that there must be a better way, especially considering disengagements for incorrect speeds are common:

From personal experience and using FSD for over 72 percent of my driving miles since v14 was released late last year, I make Speed Profile adjustments constantly. If FSD is traveling a tad too quickly, I will scale it back, and if it’s too conservative, I’ll make it more aggressive.

I don’t complain about making the Speed Profile changes too frequently, but it would certainly be nice to have it happen less frequently. There are far too many times I am concerned about getting a ticket, even in Standard mode.

The biggest issue for me, personally, which seems to be echoed throughout the community, is the fact that Tesla’s goal is to minimize disengagements. Many drivers are stating that speed is a major reason for disengagements.

However, Tesla is not willing to bring back this one level of input because it would technically be a regression.

Whether it’s right or wrong in your opinion, it is what Tesla is going with, and it seems like it has pivoted quite a bit from its other strategies for minimizing interventions by pushing its AI to behave in a way that would fit the occupant’s personal preferences.

Continue Reading

News

Tesla qualifies for awesome new first-time EV buyer incentive in California

Published

on

White Tesla Model X rear bumper showing California license plate

Tesla is one of several automakers whose vehicles qualify for an awesome new first-time EV buyer incentive program in California.

The Golden State launched the MyFirstEV incentive program, which helps those buying an electric vehicle for the first time with a $3,500 incentive on new-inventory purchases of a Model 3 or Model Y.

The incentive requires an order on or after August 3, and delivery must be taken while the program is still being funded. California has set aside $135.5 million to help strengthen its SEV market and support automotive innovation.

Incentives are offered at the point of sale, and used EVs are also available for a partial incentive of $1,750. Half of the $3,500 and $1,750 incentive amounts are covered by California, with the other half being covered by participating OEMs.

Additionally, rules apply for MSRP and how the vehicle will qualify for the incentive. Any vehicle from a non-California headquartered OEM must have an MSRP of $50,000 or less. Used vehicles must be priced at $25,000 or less and must be at least two model years older than the year of purchase.

The cars must also be purchased from manufacturers as certified pre-owned vehicles. Private dealerships are not eligible.

In total, California expects to incentivize over 73,000 ZEVs.

Participating Manufacturers

Fourteen total automakers are participating in California’s MyFirstEV program:

  • Chevrolet – Launching August 2026
  • Ford – Launching August 2026
  • Honda – Launching September 2026
  • Hyundai – Launching August 2026
  • Kia – Launching August 2026
  • Lexus – Launching September 2026
  • Lucid – Launching August 2026
  • Mitsubishi – Launching November 2026
  • Nissan – Coming Soon
  • Rivian – Coming Soon
  • Subaru – Launching September 2026
  • Tesla – Launching August 2026
  • Toyota – Launching September 2026
  • Volvo – Coming Soon

 

Continue Reading

News

Tesla Full Self-Driving insurance program with heavy discount expands

Published

on

Lemonade has expanded its innovative Autonomous Car insurance program to Tennessee, giving Tesla owners in the state a substantial discount on Full Self-Driving (FSD) miles. Announced on August 3, the product offers 50 percent off every mile driven with FSD activated, positioning the digital insurer as a leader in pricing insurance around autonomous technology.

The program, marketed as Lemonade Autonomous Car insurance, uses a direct connection via Tesla’s Fleet API (with customer permission) to automatically distinguish FSD-engaged miles from manual driving. Policyholders pay a low base rate when the vehicle is stationary and a few cents per mile when moving, with the 50 percent reduction applied specifically to FSD miles.

Coverage includes standard protections such as liability, collision, comprehensive, roadside assistance, and Tesla-specific benefits like access to certified repair shops and emergency crash services. Eligible vehicles require Hardware 4, as well as recent firmware.

Lemonade first unveiled the product on January 21 of this year, describing it as a first-of-its-kind offering designed for self-driving cars, starting with Tesla FSD. It began rolling out in Arizona on January 26, followed by Oregon about a month later. Subsequent expansions brought it to Indiana in early June 2026 and Colorado later that month.

Tennessee marks the fifth state.

Tesla Full Self-Driving gets outrageous insurance offer with insanely cheap rates

The discount rests on Lemonade’s strong belief in the safety of Tesla’s FSD system. The company cites Tesla’s data showing that FSD-driven miles are twice as safe as those driven manually, or associated with roughly a 50 percent crash reduction.

Lemonade Co-founder and President Shai Wininger has emphasized this distinction: “Traditional insurers treat a Tesla like any other car, and AI like any other driver. But a car that sees 360 degrees, never gets drowsy, and reacts in milliseconds can’t be compared to a human.”

He added that “Teslas driven with FSD are involved in far fewer accidents” and committed that as FSD software improves and becomes safer, Lemonade’s prices will drop further.

Tesla Full Self-Driving gets an offer to be insured for ‘almost free’

This approach leverages Lemonade’s existing pay-per-mile technology and AI-driven risk models, which analyze nuanced vehicle data including software version and sensor performance. The company expects the model to reward higher FSD usage with greater savings while supporting mixed households that include both Tesla and non-Tesla vehicles under one policy. Bundling with home, renters, or pet insurance can yield additional discounts.

As autonomous driving technology advances, Lemonade’s state-by-state expansion of usage-based pricing that directly reflects real-world safety data represents a notable shift in how insurers evaluate risk.

Tesla owners in the five available states – Arizona, Oregon, Indiana, Colorado, and now Tennessee – can obtain quotes quickly through the Lemonade app or website, potentially lowering the overall cost of ownership for vehicles equipped with advanced driver-assistance systems. Further states are expected as regulatory approvals progress.

Continue Reading