Tesla Earnings Call Was More Complicated Than Expected

Tesla’s latest update signals that its autonomous vehicle strategy is moving into high gear, with Cybercab production now underway at Giga Texas, volume output targeted for 2026, and plans to eventually make it the company’s highest-volume vehicle. FSD usage has doubled to nearly 29 million miles per day, adoption is accelerating across the fleet, and the subscription business is becoming a high-margin cash flywheel. Musk expects robotaxis to operate in about a dozen states by year-end, though he stresses that global scale requires another major software version and that near-term revenue will be modest. For Hardware 3 owners, Tesla will offer both discounted trade-ins and low-cost upgrades to AI4 hardware, potentially around $800 with free FSD, using temporary “micro factories” to make the conversions efficient—while FSD Version 14 Light, slated for June, will bring point-to-point functionality to older vehicles. Overall, the message is that Tesla is carefully balancing aggressive robotaxi expansion with hardware transition realities, but sees FSD and Cybercab as the dominant drivers of future growth.

• ▶ 0:13 Tesla announced the Cybercab has started production at Giga Texas, with volume production expected this year and plans to eventually replace the Model Y as its highest-volume vehicle.
• ▶ 0:29 FSD usage has doubled to 29 million miles per day, and Elon Musk confirmed FSD version 15 could arrive on AI4 cars by the end of the year.
• ▶ 0:47 Musk expects the robotaxi to operate in a dozen or so states by year-end, shifting the framing from safety to operations—sparking a debate over whether wider rollout is being delayed or refers to global unsupervised FSD.
• ▶ 2:08 Musk confirms Tesla has just started production of the Cybercab and will begin Semi production soon, warning that new supply chains will follow a slow, stretched-out S-curve before ramping up exponentially.
• ▶ 3:14 Brian says the Semi bottleneck is that production hasn’t started and key pieces are still missing; he flags the 4680 battery line as a key constraint since each Semi uses the equivalent of 10–12 car batteries.
• ▶ 4:39 Tesla’s deck confirms Cybercab, Semi, and Megapack 3 are on schedule for volume production starting in 2026, and that Cybercab is expected to replace Model Y and become Tesla’s largest vehicle over time.
• ▶ 6:38 Tesla is targeting unsupervised FSD/robotaxi in a dozen or so states by end of this year, with a very cautious approach and no injuries or fatalities so far; robotaxi revenue won't be “super material” this year but could be significant next year.
• ▶ 8:02 Even if robotaxi revenue comes in 2025, it likely won't offset the infrastructure and supervision costs already invested; the real financial impact is expected next year, and hitting the originally announced 2025 targets by late 2026 would be fantastic progress.
• ▶ 10:03 Musk’s comments on Version 14.3 being the “last piece of the puzzle” do not mean robotaxi won't launch at all this year — it means 14.3 is the last piece for initial deployment, while another version with major architectural improvements is needed before very large-scale rollout.

• ▶ 15:08 FSD adoption is accelerating: fleet drives 28.8M miles/day on FSD, double the rate from a couple months earlier, and incremental FSD take rate is over 50% excluding free trials.

• ▶ 16:14 FSD is becoming a major cash flywheel: net +170K subscriptions last quarter adds ~$50M/quarter in revenue, with total FSD subscription revenue estimated at ~$375M/quarter in essentially pure margin.

• ▶ 13:38 On the robo-taxi timeline, Musk referenced “12 states,” highlighted that another software version is needed before global scale, and suggested the path may jump from 14.3 to version 15 — while the real holdup is handling edge cases, not safety.

• ▶ 19:15 Hardware 3 cannot achieve unsupervised FSD, which is the core reason Tesla is offering alternatives.
• ▶ 19:20 Tesla offers two paths for Hardware 3 owners: a discounted trade-in toward an AI4 car, or an upgrade replacing both the computer and cameras to Hardware 4.
• ▶ 20:14 Musk says efficient upgrades will require setting up "micro factories" in major metro areas, and the long-term goal is converting all Hardware 3 cars to Hardware 4 for robotaxi fleet eligibility.
• ▶ 21:03 Upgrading Hardware 3 at service centers would be "a pain" and cost prohibitive, but a temporary assembly-line facility running the same changeover repeatedly would be efficient and cheap.
• ▶ 21:24 The upgrade process alone could cost as low as a few hundred dollars at cost; including AI4 hardware it might reach around a thousand dollars, and since the chip has been in production for years, the overall upgrade is likely cost effective.
• ▶ 21:45 Tesla may need special pricing for FSD subscribers versus owners, such as factoring in how long they've paid or offering a credit, meaning upgrade costs might not be uniform.
• ▶ 21:52 Tesla may offer six months of free FSD as an incentive for Hardware 3 upgrades, though the exact plan is still unclear even to the hosts.
• ▶ 22:17 FSD transfer rights are not fully confirmed; a discount on a new purchase may be the intended mechanism, and the incentive likely applies only to those who bought FSD outright, not subscribers.
• ▶ 22:52 Because robotaxi ambitions benefit from more capable cars, Tesla may offer a low-cost upgrade path to a broader group of owners—potentially around $800 with free FSD included.
• ▶ 23:15 The host reacts positively to news about FSD Version 14 Light, calling it “really good news,” though he admits he hasn't independently verified it yet.
• ▶ 23:26 The company confirmed there will still be FSD Version 14 Light, despite other Hardware 3-related discussions.
• ▶ 23:28 The expected release window is stated as June, with the host repeating “He said June, right? June.”
• ▶ 23:32 By June, all Hardware 3 cars are expected to receive FSD 14 Light, a company-committed rollout window.
• ▶ 23:40 The update applies to older, existing HW3 vehicles like the speaker's daughter's Model 3, not just new cars.
• ▶ 23:47 Per Tesla's Ashok, FSD 14 Light is "basically point-to-point" and should retain most of the core functionality of the full v14 system.
• ▶ 23:56 Hardware 3 owners get two options: trade in their car for AI4 hardware, or have Tesla replace the hardware directly.
• ▶ 24:06 Tesla kept its promise to Hardware 3 buyers, but the upgrade took years—and both trade-in and hardware replacement paths will be offered.
• ▶ 24:19 FSD 14 Light is the supervised option for owners who don't want unsupervised driving, while the core goal is making Hardware 3 cars robotaxi-capable so owners can make money from them.
• ▶ 25:07 AI5 won't go into cars soon because AI4 is expected to achieve unsupervised FSD – called "the most important thing" for Tesla's near-term roadmap.
• ▶ 25:26 AI4 alone will hit the safety levels needed for unsupervised operation, so Tesla doesn't need AI5 or new hardware to reach FSD safety targets.
• ▶ 25:34 AI4 Plus details: more RAM (16 to 32 GB per chip, interpreted as 64 GB dual), ~10% compute increase, production starting mid-2027, and dependent on Samsung's ability to make the chip.
• ▶ 25:53 The speaker confirms the call validated known points and filled in gaps on unknown ones, giving a content-based positive read.
• ▶ 26:27 Reading only the transcript led to a “bright” impression, because they were not exposed to the live call’s tone of voice.
• ▶ 26:34 The speaker explicitly notes they avoided the “tone of voice penalty,” highlighting how written transcripts can make an earnings call seem more positive than hearing it live.
• ▶ 26:37 Panel unanimously identified robotaxi rollout as the single most important metric before the earnings call.
• ▶ 26:56 Despite the call's negative "downer" tone, the speaker notes there was no delay announcement—only details like 12 states, FSD 14.3 not going wide immediately, and a subsequent version expected to go wide.
• ▶ 27:28 With Cybercab production ramping from hundreds to thousands per week, a wide robotaxi rollout need not wait for FSD 15; the entire debate hinges on your "definition of scale."
• ▶ 28:10 Speakers doubt what Elon Musk meant by "wide," joking it could be as limited as "Cheyenne, Wyoming" — reinforcing that "wide" may be deliberately vague.
• ▶ 28:27 One interpretation: FSD version 14.3 "is not good enough," explaining why only small-scale tests are being seen; a new version is needed before proper scaling.
• ▶ 28:40 The segment lays out competing rollout explanations: either a new FSD version is required before the 12 states/cities, or the 12 states come first and a wider expansion (possibly FSD version 15) follows later — with no firm conclusion reached at ▶ 28:59.
• ▶ 28:59 The central takeaway: FSD, CyberCab, and robotaxi are the most important areas for Tesla investors to follow.
• ▶ 29:20 Viewers are directed to follow Brian at his channel, Futurasa.
• ▶ 29:26 The host promotes herbalm.com as a comprehensive resource for Tesla investors.

• ▶ 0:00 Intro and Initial Reaction (0:00 - 2:02) - - Hosts react to the Tesla earnings call and introduce Tesla journalist Brian White.
• ▶ 2:02 Cybercab and Semi Production (2:02 - 6:38) - - Brian analyzes Cybercab and Semi production ramp, battery constraints, and Tesla's autonomous future.
• ▶ 6:38 Robotaxi and Unsupervised FSD Outlook (6:38 - 12:34) - - They debate robotaxi rollout, 2026 targets, and Elon's 14.3 unsupervised FSD comments.
• ▶ 12:34 FSD Adoption, Tailwinds, and Software Roadmap (12:34 - 19:03) - - Discussion moves to inventory tailwinds, FSD take rates, Version 15, and AI hardware expectations.
• ▶ 19:03 Hardware 3 Upgrades and Final Takeaways (19:03 - 29:35) - - They cover the Hardware 3 upgrade path, AI4 trade-ins, and wrap up the earnings call.