Elon Musk's electric vehicle maker Tesla has announced plans to introduce unsupervised Full Self-Driving (FSD) capability to customer vehicles as early as the second quarter of next year.
Ashok Elluswamy, Tesla's vice president of AI software, stated during an earnings conference call for the third quarter held on the 21st (local time) that "by around the second quarter of next year, we expect to operate unsupervised FSD on some customer-owned vehicles so they can be included in the robotaxi fleet."
Currently, Tesla's FSD is a supervised system where drivers must keep their eyes on the road and respond to unexpected situations, but this restriction is expected to be lifted soon.
Under this setup, Tesla owners will be able to generate revenue by picking up passengers like a taxi when they are not using their vehicles.
Chief Executive Officer Elon Musk also noted that demand for Tesla vehicles is surging, stating, "The biggest driver of our vehicle demand growth is FSD."
He expressed confidence, adding, "No car available in the market today competes with us in terms of FSD."
During the third quarter in North America, more than 75% of new Tesla vehicle deliveries opted for paid FSD subscriptions, a 20 percentage point increase from the previous quarter, according to the company.
The number of active FSD subscribers worldwide also reached 1.74 million, up 67% from the previous year.
Regarding the autonomous "robotaxi" service, CEO Musk emphasized that "the number of cybercabs in Austin (Texas) has increased tenfold since their launch a month ago," and described the cumulative mileage on weekly charts as "mind-blowing."
Cumulative mileage for Tesla robotaxis approached 2 million miles (approx. 3.2 million km) as of September this year.
Elluswamy previewed that the robotaxi service will expand to more cities, launching in Dallas next month.
Additionally, he unveiled a schedule for Terafab, the mega semiconductor plant under construction in Austin, Texas, aiming to begin equipment test runs in the first half of next year, produce test chips in May and June, and start commercial production from the third quarter of 2028.
CEO Musk remarked, "Terafab will be the most valuable building on earth," adding that its economic value will rival the top 20 economies in the world.
He explained that capital expenditures (CapEx) running into tens of billions of dollars will be made through early 2028, with long-term spending expected to exceed $100 billion (approx. 130 trillion won).
Despite surging expenditures, Tesla reported on this day that its third-quarter revenue reached $28.05 billion (approx. 37.2 trillion won), beating the market consensus of $27.9 billion compiled by the London Stock Exchange Group (LSEG).
Earnings per share (EPS) also came in at 48 cents, higher than the market estimate of 43 cents.
By sector, revenue from the core automotive business fell 3% year-over-year to $20.56 billion, and the energy generation and storage sector saw revenue drop 16% over the same period to $2.88 billion.
Conversely, the services and other segment—spanning charging, insurance, and repairs—surged 33% from the same period last year to generate $4.61 billion, buffering the overall revenue decline.
Total vehicle deliveries in the third quarter dipped 2% year-over-year to 486,532 units, though select countries including Canada, Taiwan, Japan, and Australia achieved record quarterly deliveries.
Due to heavy investments in artificial intelligence (AI) and robotics, third-quarter capital expenditures (CapEx) more than doubled from the same period last year ($2.248 billion) to reach $4.732 billion.
Consequently, increased cash outflows led to a 62% plunge in third-quarter free cash flow (FCF) to $1.531 billion compared to a year earlier.
However, compared to the previous quarter's deficit of approximately $1.1 billion, the company successfully swung back to a profit within a single quarter.
Nevertheless, this level of capital expenditure remains significantly lower than market analysts' expectations of $7.5 billion.
(Photo: Getty Images)
※
Copying, redistribution, and unauthorized use in AI training are strictly prohibited.