Top News

• Waymo said cameras alone are insufficient, defending its use of lidar, radar and HD maps.
• Ferragu accused Waymo of “cognitive dissonance” and said its mapping dependence reflects legacy architecture.
• He argued that improving vision AI could make Waymo’s sensor-heavy “driving gas plant” increasingly irrelevant. 

Shares of Tesla, Inc. (TSLA) are on track to snap a three-week winning streak, but New Street Research analyst Pierre Ferragu said Alphabet, Inc.-owned Waymo’s defense of its sensor-heavy robotaxi system made him more confident in Tesla’s vision-based approach.

TSLA stock jumped 3% on Thursday, but shares are still down 2% for the week. 

Waymo Defends Sensor-Heavy Approach 

“Love it. Waymo getting defensive. Nothing makes me more bullish about Tesla,” Ferragu said on X. “Arguments they bring forward are poor.” New Street has a ‘Buy’ rating and $600 price target on TSLA, implying a 69% upside from current levels. 

Ferragu was responding to a Waymo post outlining 10 lessons from more than 200 million fully autonomous miles. Alphabet’s robotaxi unit said that cameras alone cannot deliver safe autonomy at scale. Its vehicles combine cameras, lidar and radar for greater redundancy across different road and weather conditions.

Waymo also defended high-definition maps and fewer, larger AI models while rejecting fully end-to-end systems that convert raw visual data directly into driving commands. “You can’t build trust with a black box,” Waymo said, arguing that a separate safety layer should validate proposed driving paths against physical constraints and traffic laws.

The company called the idea that supervised Level 2 driver assistance can simply evolve into Level 4 autonomy a “false summit,” adding that there is “no substitute” for actual driverless experience.

Ferragu Flags Waymo Contradictions 

Ferragu argued that Waymo’s case contained contradictions. Waymo cited more than 200 million autonomous miles as evidence that multimodal sensing is indispensable while warning that statistics alone cannot establish trust in opaque AI systems, he said.

“See the cognitive dissonance here?” Ferragu said. “Either you accept statistics as proof of safety, or you don’t.” He also questioned why Waymo still needs HD maps if its cameras, lidar and radar already provide reliable perception in challenging conditions. Ferragu said mapping was a legacy requirement built into Waymo’s original architecture.

Ferragu said increasingly capable AI was reducing the risk that cameras would misinterpret their surroundings, which multimodal sensing was partly designed to solve. “The disturbing truth for Waymo today is that multimodal was meant to compensate for the risk of misperception of vision,” he said.

Ferragu called Waymo’s system a “driving gas plant” that AI at scale could make irrelevant, calling the company’s position “incumbent rhetoric” and “innovator dilemma 101.”

Ferragu Backs Tesla’s FSD 

Tesla relies primarily on cameras, neural networks, and data collected across its consumer fleet. The approach could offer lower costs and faster expansion if it can demonstrate adequate safety without specialized sensors and pre-mapped operating areas. Waymo, however, has accumulated more than 200 million fully autonomous miles, giving it substantially more real-world driverless experience.

Ferragu said last month that investors underestimate Tesla’s Full Self-Driving (FSD) software. He pointed to FSD net additions exceeding U.S. vehicle deliveries in the first quarter and stronger second-quarter deliveries as potential signs of an inflection.

“The vast majority of the market has no idea what FSD can do, and discovers it only progressively,” Ferragu said. He compared Tesla’s potential evolution with the iPhone, whose market initially appeared limited when consumers viewed it only as an expensive phone. “A Tesla is not a car, the same way an iPhone was not a phone,” Ferragu said.

How Do Retail Traders Feel About TSLA?

On Stocktwits, retail sentiment slipped to ‘neutral’ from ‘bullish’ levels a day ago amid ‘normal’ message volume.

One user said, “$TSLA who believes earnings in automatic vehicles will automatically be higher than current cars? Are investors that dumb? Robots are going to be so cheap”

Another user said, “$TSLA The thing that kills me is China is having literal robot olympics over there, and we're over here with our puds in our hands waiting for the Roaster that was announced in 2006, literally 20 years ago.”

So far this year, Tesla's stock has lagged its "Magnificent Seven" peers, making it the group's worst performer, down about 21%. 

For updates and corrections, email newsroom[at]stocktwits[dot]com.<