Divergent Compute.AI Economic Think Tank

Markets / Industries / Automotive & Mobility

Industry view · Automotive & Mobility

The car learned to drive itself; the question is who collects the fare

Robotaxi crossed from demo to revenue in 2025: Waymo cleared 450,000 weekly paid rides and 14 million trips for the year, Baidu's Apollo Go matched it in China at 250,000 weekly, and Aurora put driverless heavy trucks on Texas freeways. The value is migrating from the metal to the model, the sensor stack, and the operator that owns the network.

$18.8B
Automotive AI market, 2025
MarketsandMarkets
450K
Waymo weekly paid rides
CNBC
$126B
Waymo valuation, Feb 2026
Waymo
$1.7B
Nvidia automotive FY2025 rev
Nvidia

01 · The thesis

The disruption is real, but it is bifurcating, not converging

Two business models have separated. Waymo's sensor-heavy, geofenced robotaxi is now a commercial service compounding at scale — ridership roughly doubled inside a year — and its $126 billion February 2026 valuation prices it as a category winner, not a research project. Tesla's camera-only bet went live in Austin in June 2025 but still ran a fleet of only ~20 vehicles a year later, a reminder that a software-first thesis is not the same as a deployed network.

The economics now favor whoever owns the operating network and the silicon, not the badge on the hood. Nvidia's automotive line hit $1.7 billion in FY2025 and Mobileye is sitting on 46 million EyeQ6 Lite orders; meanwhile legacy OEMs increasingly license the autonomy stack rather than build it. China is a parallel, faster-scaling universe: Apollo Go passed 17 million cumulative rides and Pony.ai and WeRide both IPO'd in Hong Kong in late 2025.

1AI silicon

The pick-and-shovel layer

Inference chips and full-stack drive computers are where the most durable margin sits.

Nvidia DRIVE, Mobileye EyeQ, Qualcomm Snapdragon Ride
2Perception & stack

From hand-coded rules to end-to-end nets

The industry is shifting to learned, end-to-end driving models that fuse camera, radar and lidar.

Waymo, Tesla FSD, Mobileye, Pony.ai
3Robotaxi ops

The network is the business

Owning the fleet, the depots and the rider app captures the recurring fare, not the OEM.

Waymo, Apollo Go, Pony.ai, WeRide
4Driver assist (L2/L3)

Autonomy that ships at volume today

Hands-free highway systems are the real near-term AI revenue inside privately owned cars.

GM Super Cruise, Mercedes Drive Pilot, Tesla
5Freight & logistics

Driverless trucks find product-market fit first

Fixed highway routes and a labor-shortage tailwind make autonomous trucking commercially live now.

Aurora, Kodiak, Waymo Via legacy
Pace of AI disruption by stage — Divergent Compute assessment

02 · Public players & exposure

Who routes through, who gets routed around

We plot the listed players on two editorial axes — how exposed each is to AI disruption, against how ready its data, brand and position are to be the answer. The figures in the table are sourced; the placement is our read.

Positioning — editorial assessment, not a sourced metric. Bubble = approximate relative scale.
CompanyStanceThe sourced fact
Waymo (Alphabet)GOOGLCategory leaderRaised $16B at a $126B valuation in Feb 2026 after crossing 450K weekly paid rides and 14M trips in 2025 (Waymo; CNBC).
NvidiaNVDAOwns the siliconAutomotive revenue rose 55% to $1.7B in FY2025 and is in full production on its DRIVE stack for the Mercedes CLA (Nvidia).
MobileyeMBLYVolume incumbentGuided to ~$1.85-1.88B 2025 revenue and holds orders for 46M EyeQ6 Lite assisted-driving chips (Mobileye; SEC).
TeslaTSLABet not yet provenLaunched Robotaxi in Austin June 2025 but still ran only ~20 vehicles a year later versus Waymo's hundreds (Electrek).
BaiduBIDUChina leaderApollo Go passed 17M cumulative rides by Oct 2025 and hit 250K weekly rides, matching Waymo's spring pace (CNBC).
Pony.aiPONYScaling pure-playQ1 2026 revenue up 145% YoY; fleet topped 1,700 robotaxis with a raised 3,500-vehicle 2026 target (Bloomberg).
AuroraAURFreight first-moverBegan the first commercial driverless heavy-truck service on public roads, Dallas-Houston, May 2025 (Aurora; SEC).
General MotorsGML2 at scaleCustomers logged 1 billion hands-free Super Cruise miles by April 2026, with 500K+ equipped vehicles on the road (GM).
Mercedes-BenzMBGL3 pioneerFirst certified SAE Level 3 system in the US, approved in California and Nevada on mapped freeways (Mercedes-Benz).
WeRideWRDCash-burn watchHong Kong IPO raised HK$2.4B for the still-unprofitable firm to fund autonomous-driving R&D (Yahoo Finance).
The map is Divergent Compute’s editorial positioning, offered as a lens, not a measurement. Every figure in the right-hand column is drawn from a named source — see Sources.

03 · The two clocks

The spend, and the payoff

Three timers running at very different speeds across the industry

Source: CNBC, Waymo, GM, Nvidia, Mobileye company disclosures (2025-2026)

Robotaxi scaling is the fast clock. Waymo roughly doubled weekly ridership inside a year to 450,000 paid rides and is targeting 1 million per week by end-2026. This is the surface where AI is already converting to recurring revenue.

Driver-assist adoption is the steady clock. GM crossed 1 billion cumulative hands-free Super Cruise miles by April 2026 with 500,000+ equipped vehicles, while Mercedes holds the only certified Level 3 approval in the US. Volume autonomy is shipping inside cars people already own.

The China clock runs in parallel and fast. Apollo Go logged 17 million+ cumulative rides and 3.4 million in Q4 2025 alone, up over 200% YoY, while Pony.ai and WeRide tapped public markets to fund fleets across China and the Gulf.

04 · Private flagships

The AI-native challengers

The companies attacking this industry AI-first, with disclosed funding where available:

Waymo

Robotaxi network operator

The clear commercial leader: 14M paid trips in 2025, expanding city by city with a sensor-rich, mapped approach.

$16B round at $126B post-money valuation, Feb 2026 (Waymo)

Pony.ai

China robotaxi pure-play

Toyota-partnered fleet scaling past 1,700 vehicles toward a 3,500-car 2026 target, with revenue up 145% YoY.

~$3.9B market cap; Hong Kong + Nasdaq listed (Bloomberg, May 2026)

Aurora Innovation

Autonomous freight

First driverless commercial heavy-truck operator on public roads, running the Dallas-Houston corridor.

Public (AUR); commercial launch May 2025 (SEC)

WeRide

Multi-market AV developer

Deploying robotaxis and robobuses across China and the Middle East; still pre-profit and consuming cash.

HK$2.4B raised in Nov 2025 Hong Kong IPO (Yahoo Finance)

Mobileye

Assisted-driving silicon

The volume backbone of OEM ADAS, with a 10M-unit Volkswagen order and 46M EyeQ6 Lite chips on order.

Public (MBLY); ~$1.87B guided 2025 revenue (SEC)

05 · Signals

What moved, and what to watch

May 2025

Driverless trucks go commercial

Aurora begins the first paid driverless heavy-truck deliveries on public roads between Dallas and Houston.

Jun 2025

Tesla Robotaxi goes live in Austin

Tesla launches its camera-only robotaxi pilot with safety monitors, beginning its long-promised network.

Nov 2025

China's AV pure-plays hit public markets

Pony.ai and WeRide complete Hong Kong IPOs as Apollo Go matches Waymo's spring ridership pace.

Feb 2026

Waymo prices as a winner

A $16B raise values Waymo at $126B, nearly triple its October 2024 mark, on commercial ridership growth.

Apr 2026

Driver assist crosses a billion miles

GM reports 1 billion cumulative hands-free Super Cruise miles, proving L2 autonomy at consumer scale.

06 · The exposure read

Who’s defensible, who’s at risk

AI rewards clean, structured advantage and punishes friction. The line runs through who owns the data, the brand and the customer — and who is merely a step the technology can route around.

Defensible

  • Silicon and stack suppliers — Nvidia and Mobileye monetize autonomy regardless of which OEM or operator wins, selling the compute and perception underneath every approach.
  • Network operators with a head start — Waymo and Apollo Go compound ridership and data faster than entrants can catch, and capital markets are now pricing that lead explicitly.
  • Freight specialists on fixed routes — Aurora and peers found product-market fit first, where mapped highway corridors and a driver shortage make the unit economics work today.
  • Premium OEMs that license, not build — Mercedes and others shipping certified L3 and partnering on the stack capture autonomy value without funding a full AV moonshot.

At risk

  • Sub-scale robotaxi pure-plays — operators burning cash without a path to dense, profitable networks face a brutal capital market; WeRide remains unprofitable post-IPO.
  • OEMs that own neither the silicon nor the network — traditional automakers risk being reduced to commodity hardware vendors as value migrates to the AI stack and the fare-collecting operator.
  • Camera-only bets that can't scale operations — a software thesis is not a deployed fleet; Tesla's ~20-vehicle Austin footprint a year in shows deployment, not demos, is the bottleneck.
  • Professional drivers in freight and ride-hail — the labor displacement that makes the unit economics attractive is also the clearest human cost as driverless miles compound.
The autonomy debate has moved past whether the car can drive itself — Waymo, Apollo Go and Aurora settled that with paid miles. The open question for 2026 is industrial: who can scale a profitable network, who owns the silicon underneath it, and which legacy nameplates avoid being commoditized into the chassis. Figures cited are company and analyst disclosures as of mid-2026 and will move; the structural direction will not.

Sources

Where this comes from

MarketsandMarkets — Automotive AI market ($18.83B 2025)  ·  CNBC — Waymo crosses 450,000 weekly paid rides  ·  Waymo — $16B round at $126B valuation  ·  Nvidia — Q4 and FY2025 financial results (auto $1.7B)  ·  Electrek — Tesla Robotaxi expands Austin metro, ~20 vehicles  ·  CNBC — Baidu robotaxi rides hit 250,000 weekly  ·  Mobileye — Q3 2025 results and 2025 outlook (SEC)  ·  Bloomberg — Pony AI lifts 2026 robotaxi fleet goal to 3,500  ·  Aurora — Begins commercial driverless trucking in Texas  ·  GM — 1 billion hands-free Super Cruise miles  ·  Mercedes-Benz — Drive Pilot, first US-certified SAE Level 3  ·  Yahoo Finance — Pony.ai and WeRide Hong Kong IPO debut  ·  Forbes — Waymo targets 1 million robotaxi rides a week