TechNewsFirst

Tech news for people who actually use the tech.

Tesla's $10 Trillion Bet in 10 Years

Full Self-Driving went from an $8,000 one-time purchase to a $99-a-month rental nobody can buy outright anymore, after peaking at $15,000, years of marketing that didn't hold up, and now enough real driverless miles on Austin's streets that Wall Street is arguing Tesla's self-driving network could be worth more than its cars. That's not a footnote. It's the same subscription move software, gaming and Amazon Prime made before it, this time paired with a technology that finally started working.

Editorial illustration for: Tesla's $10 Trillion Bet in 10 Years
SourcesNotesELI5AITIArticle

What was checked, and what could not be.

This is a revised, expanded version of an earlier draft, incorporating additional research requested after the first pass: the $15,000 FSD price peak, Musk's 'appreciating asset' marketing claim and its failure, and sourced figures for the gaming-microtransaction, App Store-subscription, Amazon Prime and self-storage comparisons that were previously only gestured at as general context.

FSD pricing: the $12,000 (January 2022) and $15,000 (announced August 21, 2022 for a September 5, 2022 effective date) figures are sourced to TechCrunch's contemporaneous reporting, which also confirms Musk announced the increase himself. The subsequent cuts (back to $12,000 in September 2023, to $8,000 in April 2024, subscription from $199 to $99/month in April 2024) and the February 14, 2026 elimination of the outright-purchase option are consistent with the Not a Tesla App and CarCody pricing trackers and DriveNewsNetwork coverage used in the earlier draft.

The 'appreciating asset' claim is sourced to two Electrek pieces: the original April 2019 reporting on the Lex Fridman podcast quote, and Electrek's own March 2025 follow-up assessing the claim against what actually happened (used Teslas depreciating faster than average, FSD's price falling rather than rising since 2022). The JerryRigEverything remark and the 'roughly a fifth of its original value' figure are cited as reported via that Electrek piece and related coverage; this outlet did not independently verify JerryRigEverything's own valuation methodology and is treating that specific figure as an attributed claim, not an audited one.

Take-Two/GTA figures: the load-bearing number here, recurrent consumer spending equaling 78.1% (roughly $4.1 billion) of $5.35 billion in FY2023 net revenue, comes from Take-Two's own SEC filing (Form 8-K, fiscal Q2 2023 earnings release), an audited, company-disclosed figure. The separate '$5 billion in Shark Cards since launch' and '4% of players' figures are explicitly flagged in the piece as originating from a 2025 data breach, not an official Take-Two disclosure, and are presented as directionally consistent with, rather than equal in reliability to, the SEC-filed number.

Mobile/App Store in-app-purchase and subscription figures ($82 billion in 2025 mobile-game IAP revenue, roughly 44% of iOS App Store revenue from subscriptions) are sourced to industry-tracking aggregators (BusinessOfApps and similar market-research compilations) rather than to any single company's own disclosed financials; this is disclosed explicitly in the piece as a lower tier of sourcing reliability than the SEC-filed Take-Two number.

Amazon Prime's price history ($79 in 2005, $99 in 2014, $119 in 2018, $139 in 2022) is sourced to CNBC's contemporaneous February 2022 reporting on the most recent increase, cross-checked against a historical timeline compiled by industry analytics coverage for the earlier years; no confirmed increase beyond $139/year was found for 2023 through 2026 at the time of writing.

The self-storage 'Existing Customer Rate Increase' (ECRI) practice, including the twice-a-year cadence and the 10-20% additional cumulative revenue estimate versus one annual increase, is sourced to Inside Self Storage, a trade publication covering the industry, which discusses the practice as a named, openly discussed strategy rather than a hidden or alleged one.

Deliberately did not repeat the commissioning brief's own recollected dollar figures where they didn't match verified numbers; the actual sourced figures and dates, including the $15,000 peak the brief flagged as missing from the first draft, are used throughout instead. The '$10 trillion market cap in 10 years' framing is presented explicitly as a labeled, dated prediction (dated September 19, 2026, scored against Tesla's market cap in September 2036), not as an assertion of fact, per house rules on separating scoreable predictions from unfalsifiable claims. The firsthand Model X ownership anecdote is included as the outlet's own practitioner experience and is not attributed to any named private individual.

This version adds a further round of research, specifically requested to establish whether FSD's real-world driving capability supports treating $99/month as good value, and to connect the pricing shift to Wall Street's separate debate over Tesla's valuation as an 'AI platform' rather than an automaker. The FSD v14.3.2-14.3.4 software-unification timeline and the Forbes 'Level 4 driverless ADAS' quote are sourced to Brooke Crothers' Forbes piece, published June 14, 2026 ('Tested: Tesla FSD Is Evolving Into A De Facto Robotaxi'). The Austin robotaxi expansion figures (245 square miles reached June 3, 2026; expansion to Houston, Dallas, Miami, Orlando and Tampa by July 2026) are sourced to TechTimes and Tesery's contemporaneous coverage. Tesla's own 380,000-unsupervised-mile and nearly-1.5-million-paid-FSD-customer disclosures are sourced to Tesla's Q2 2026 earnings call as summarized by Not a Tesla App. The critical counterweight, Electrek's 19%-availability and nine-times-worse-than-human crash-rate figures, comes from Electrek's own February 16, 2026 status check, deliberately included so this piece isn't only citing the flattering numbers. The 14-versus-51 Austin crash comparison between Tesla and Waymo is sourced to NHTSA-filing coverage syndicated via Yahoo/Bloomberg; other outlets report different crash counts for different, narrower time windows, so this is disclosed as one snapshot rather than a definitive running tally. Earlier drafts of this section considered citing more specific 'miles per disengagement' and 'miles per incident' statistics from independent crowdsourced trackers; those figures could not be tied to a single verifiable, citable report in this round of research and were deliberately left out rather than published as unverified, per house policy.

The valuation figures: Morgan Stanley's 'AI platform company' reclassification and its $47/$330 per-share segment breakdown, along with its 1,500-to-30,000 robotaxi fleet growth forecast, are sourced to financial-analysis coverage (TMGM) summarizing the bank's March 18, 2026 research note, which is not itself publicly available; the same $47/$330 figures were corroborated across multiple independent financial-news aggregators. Bank of America's $750 billion robotaxi valuation is sourced to Barchart's coverage of BofA's note. Wedbush's Dan Ives $2 trillion/$3 trillion figures are sourced to his December 2025 public statement as reported by Benzinga. The bear-case figures, 2025's revenue decline to $94.8 billion, the 8.6% delivery decline, BYD's 2025 volume lead (with Tesla's Q1 2026 reclaim also noted for fairness), the ~330x trailing-earnings multiple, GLJ Research's $25.28 target and Sell rating, and the ~$407 Wall Street consensus target, are sourced respectively to Yahoo Finance's delivery/revenue reporting and Foreign Policy Journal's and MarketBeat's coverage of GLJ Research's March 2026 note. These are included specifically so the bull case above isn't presented without its counterweight.

The $379-a-month lease figure added in this revision, paired with the $99-a-month FSD subscription, is sourced to Carscoops' July 2026 piece ('Tesla Quietly Raised Its Cheapest Model 3 Lease By 12%'), which reported the base Model 3 RWD lease resetting to $379/month for 36 months (10,000 mi/year), up from $329 the week before, first flagged publicly by Sawyer Merritt on X on July 21, 2026, and corroborated by CarsDirect's contemporaneous reporting, which put the same $379 figure against a $38,380 MSRP with $4,075 due at signing. Both outlets separately calculated an effective monthly rate closer to $481 once that due-at-signing amount is amortized across the lease; this piece discloses that effective rate alongside the $379 headline figure rather than only the more dramatic-sounding number. Because Tesla's Model 3 lease pricing has swung repeatedly through 2026, cut to $299/month by CarsDirect's later reporting, then back up to $419/month and higher by September per CarsDirect's own current listings, this piece treats $379 explicitly as a dated July 2026 snapshot rather than a current, fixed price, per house policy on not presenting a volatile number as a stable fact.