The Great Self-Driving Race
Who's Leading in 2025?
Welcome back to Prdctnomics, the economics of technology products newsletter that's been on a much longer hiatus than intended. It's been a while - most of my writing was published back in 2021, and now we find ourselves in 2025 with the tech landscape dramatically transformed. Today I'm diving back in with a look at perhaps the most consequential technology megatrends happening right now: autonomous vehicles. Thanks for sticking around, and let's jump right in.
The Great Self-Driving Race: Who's Leading in 2025?
We're in the middle of one of the most expensive technology races in history.
Companies have poured over $100 billion into figuring out how to make cars drive themselves¹. And the prize? A market projected to be worth nearly $1 trillion by 2040².
But here's what most people don't realize: there isn't just one race. There are three separate battlegrounds with different leaders in each.
The Three-Way Battle for Autonomous Dominance
Battleground #1: Robotaxis
Current Leader: Waymo
Waymo - The future is now (Video)
Waymo is absolutely crushing it. They're completing over 200,000 paid driverless rides EVERY WEEK across Phoenix, San Francisco, LA, and Austin³. That's not testing - that's real people paying real money to be driven around by robots.
Alphabet just poured another $5.6 billion into Waymo⁴. They're doubling down on their investment as competitor Cruise (GM's robotaxi effort) suspended operations following a serious safety incident that led to regulatory action⁵.
I have used Waymo in San Fransisco several times now and mostly loved the experience. I personally would not use a human driver via ride sharing anywhere Waymo is available, it is that good. Waymo drove significantly better than any of my recent human Uber drivers, had cleaner vehicles, and was less expensive. This observation will become a consensus view I believe as more people try robotaxis and they continue to improve.
In China, Baidu's Apollo Go is the dominant player, with over 1.1 million rides in Q4 2024 alone⁶. They're expanding aggressively into Dubai and Abu Dhabi⁷.
As of April 2025, Baidu's Apollo Go robotaxi service operates a fleet of over 400 fully driverless vehicles in Wuhan, with plans to expand to 1,000 vehicles by the end of the year. In Dubai, the service plans to deploy 100 fully autonomous vehicles by the end of 2025, with an expansion goal of at least 1,000 vehicles by 2028. In Abu Dhabi, initial trials have begun, with full commercial operations expected by 2026.
Meanwhile, Zoox (Amazon's $1.2 billion bet) is preparing to launch its first commercial service later this year⁸. Their vehicle looks like a living room on wheels - no steering wheel, no front or back, just seats facing each other. Details on their exact fleet size are hard to come by.
And Tesla? Their plan is robotaxis in Austin by June 2025⁹ (build on modified Model Y vehicles with FSD). But we've heard "robotaxis next year" since 2019, so... ¯_(ツ)_/¯
As of April 2025, Tesla's Cybercab: a fully autonomous, purpose-built robotaxi, is not yet operating on public roads, but mass production planned to start in 2026.
Battleground #2: Autonomous Trucking
Co-Leaders: Aurora and Kodiak
While everyone's distracted by robotaxis, the trucking industry is poised to get disrupted.
Aurora Innovation is set to launch driverless commercial trucking between Dallas and Houston in April 2025¹⁰. Think about that - next month, 18-wheelers with no humans will be delivering goods on Texas highways.
Kodiak Robotics has already logged over 750 hours of completely driverless commercial operations in West Texas¹¹. Atlas Energy Solutions has ordered 100 autonomous trucks from them¹². Let that sink in - a real company is buying 100 trucks with no drivers.
Torc Robotics (owned by Daimler Truck) is targeting 2027¹³. They're taking longer, but when the world's largest truck manufacturer puts their weight behind something, pay attention.
Battleground #3: The Tech Platforms
Co-Leaders: Nvidia and Mobileye
The companies making the brains behind self-driving have the best business model of all - they sell pickaxes in a gold rush.
Nvidia dominates computing hardware with their DRIVE platform (now upgrading to their "Thor" chip based on Blackwell architecture)¹⁴. They're the arms dealer selling to everyone in the war.
Mobileye has deep relationships with nearly every carmaker on the planet¹⁵. They've built a brilliant ladder of products from basic driver assistance to full autonomy¹⁶.
Qualcomm is the challenger, leveraging their smartphone chip dominance to move into automotive¹⁷. Late to the party, but catching up fast.
The Great Technological Fork in the Road
There's a massive philosophical divide in how to make cars drive themselves.
Almost everyone (Waymo, Zoox, Aurora, Mobileye, etc.) believes you need a combination of cameras, radar, AND LiDAR (those spinning things on top of self-driving cars that cost $10,000+ each)¹⁸. This multi-sensor approach creates redundancy and safety.
Then there's Tesla.
Tesla is the only major player betting everything on cameras alone - no LiDAR, not even radar anymore¹⁹. Elon's reasoning? "Humans drive with their eyes, so cars should too"²⁰.
It's like bringing a knife to a gunfight... but if the knife turns out to be Excalibur, Tesla wins everything.
Existential Threat to Traditional Automakers
Clayton Christensen famously described three types of innovation:
Efficiency Innovation
Sustaining Innovation
Disruptive Innovation.
This framework explains not just why traditional automakers are struggling with self-driving tech, but why their entire business model faces existential danger.
Traditional automakers are masters of Efficiency Innovation - squeezing more profit from existing processes. Look at Toyota's relentless lean manufacturing or Ford's automated assembly lines. They're obsessed with reducing the cost to produce largely the same product. Even their EV strategies reflect this mindset - "How do we efficiently make electric versions of our existing lineup?"
They also excel at Sustaining Innovation - incremental improvements to existing products. Slightly better engines, marginally nicer interiors, a few more MPG each year. This worked beautifully in a world where transportation meant personal car ownership and human drivers. Their entire organizational structure, supplier relationships, dealer networks, and R&D processes are optimized for this approach.
But self-driving technology represents classic Disruptive Innovation - creating entirely new markets while making existing ones obsolete. The true disruption isn't just autonomous driving; it's what happens next:
The death of individual car ownership. When robotaxis become cheaper per mile than owning a vehicle (projected by 2035), why would consumers tie up capital in depreciating assets that sit unused 95% of the time?
A dramatic shrinking of the total vehicle market. One autonomous vehicle in constant operation could replace 5-10 personally owned vehicles. For manufacturers built to sell millions of cars to individual consumers, this means a catastrophically smaller market.
A regulatory environment increasingly hostile to human drivers. As self-driving vehicles prove statistically safer, governments will likely create incentives to get humans off the wheel. We're already seeing automotive insurance costs skyrocket; this will accelerate as human drivers become a higher-risk pool.
The complete disruption of adjacent industries - traditional auto insurance, repair shops, parking garages, and auto dealers all face massive upheaval.
Traditional automakers are responding predictably - investing primarily in efficiency innovations (cheaper EVs) and sustaining innovations (marginally better driver assistance). But they're structurally incapable of the disruptive leap needed. Their dealer networks actively fight direct-to-consumer models. Their unions resist the dramatic workforce changes autonomous tech requires. Their shareholders demand quarterly returns incompatible with the decade-long investment needed.
As Christensen observed, being great at operating in an existing paradigm often prevents you from succeeding in a disruptive one. The skills, processes, and values that made traditional automakers successful are precisely what will kill them in the autonomous age. Their expertise in building cars for human drivers to own becomes not just irrelevant but a liability in a world of robotaxis.
Strategic Approaches and Market Dynamics
The players I've mentioned are pursuing wildly different strategies:
Integrated Full-Stack Players like Waymo and Zoox are developing everything - sensors, compute, software, and in Zoox's case, even the vehicle itself. This gives them incredible control but requires massive investment.
Vision-Centric Tesla is betting it all on cameras and AI, deliberately excluding LiDAR and radar. They're leveraging their fleet data and Dojo supercomputer to train neural networks that can match human vision.
Platform Providers like Nvidia, Qualcomm, and Mobileye focus on providing the building blocks - SoCs, software development kits, and reference designs to OEMs and suppliers.
AI-First/Software Licensing is Wayve's approach - developing adaptable, data-driven AI software that can operate without HD maps, aiming for rapid scalability through licensing.
So Who Wins?
If I had to guess today:
Robotaxis: Waymo is the clear leader with unmatched scale and technology. They'll dominate the US market while Baidu owns China and broader Asia / Middle East.
Trucking: Aurora and Kodiak are neck-and-neck for the near term, but Torc/Daimler is the sleeping giant for the long haul (pun intended).
Platforms: Nvidia and Mobileye will continue to power most of the industry, with different companies choosing different partners.
The Wildcard: Tesla. If they actually solve FSD with cameras only, they could leapfrog everyone. But that's a massive IF. They may be forced to add LiDAR/radar for true L4 capability.
Why this matters
The next 24 months will be the proving ground²⁵. We'll see whether these companies can scale beyond their initial deployments and start generating actual profits.
The companies that succeed here won't just disrupt transportation; they'll fundamentally alter human civilization's relationship with distance, time, and space.
Cities will decentralize as commute time becomes productive time, transforming two-hour drives from burden to opportunity. The weekend map will expand dramatically: go to sleep in your autonomous vehicle on Friday night and wake up 500 miles away for a Saturday morning hike in the mountains. Those destinations that once required flight bookings and airport hassles become accessible with nothing more than packing a bag and falling asleep.
Meanwhile, autonomous trucking will collapse shipping costs, operating 24/7 without rest breaks, union contracts, or hours-of-service limitations. This will reshape supply chains, reduce consumer prices, and bring production closer to demand.
Above all, these vehicles will save lives, potentially millions of them. The 1.3 million annual global traffic deaths, mostly caused by human error, could plummet. The elderly and disabled will gain unprecedented mobility independence. Parents will no longer fear their teenagers driving home late at night.
In that new world, our lives will expand in ways we've only begun to imagine.
IMPORTANT DISCLAIMER
This article represents my personal opinions and analysis only, not statements of fact. All views expressed are speculative, subject to bias, and should not be construed as factual representations of reality. Content should be treated as entertainment and general information only.
I am not a registered investment advisor, financial analyst, securities broker, or financial planner. I have no professional relationship with any readers of this newsletter, and no fiduciary duty to you. Nothing in this newsletter establishes any form of advisor-client relationship.
I expressly disclaim all responsibility for any investment outcomes that may result from following ideas, concepts, or commentary presented in this newsletter. Past performance is not indicative of future results, and all investments involve risk of loss, potentially including the loss of all invested capital.
For full disclosure, I may currently hold positions (long or short) in companies mentioned in this article, including but not limited to: Alphabet/Waymo, Tesla, Nvidia, Mobileye/Intel, Qualcomm, Baidu, and Amazon/Zoox. My positions may change at any time without notice. I may have business relationships or seek to do business with companies mentioned. I may receive compensation directly or indirectly connected to these companies. These actual or potential conflicts of interest may influence my opinions despite my best efforts to remain objective.
Despite my best efforts to ensure accuracy, this article may contain errors, inaccuracies, or outdated information. The autonomous vehicle industry is evolving rapidly, with new developments occurring weekly. Data points and statistics cited were current as of publication, but may have changed by the time you read this. Company strategies, deployment timelines, and market projections are subject to change without notice.
Always consult with a qualified financial, legal, or investment professional before making investment decisions. This content should not serve as the basis for any investment actions.
References
¹ "AV Investment Trends: Who's Pouring the Most Money into Self-Driving Technology? (Funding Stats)" - https://patentpc.com/blog/av-investment-trends-whos-pouring-the-most-money-into-self-driving-technology-funding-stats
² "Autonomous Vehicle Market Size to Reach $980.7 Billion, Globally, by 2040 at 22.3% CAGR: Allied Market Research" - https://www.globenewswire.com/news-release/2025/02/04/3020166/0/en/Autonomous-Vehicle-Market-Size-to-Reach-980-7-Billion-Globally-by-2040-at-22-3-CAGR-Allied-Market-Research.html
³ "Waymo - Wikipedia" - https://en.wikipedia.org/wiki/Waymo
⁴ "How to invest in the second wave of autonomous driving" - https://globalventuring.com/corporate/asia/invest-autonomous-driving/
⁵ "NHTSA closes preliminary investigation into General Motors' Cruise robotaxis" - https://apnews.com/article/cruise-gm-nhtsa-general-motors-odi-48591349c24c6f969c07c65f1ed4e9d8
⁶ "5 Autonomous Vehicle Leaders Revolutionizing Transportation (2025)" - https://www.securities.io/autonomous-vehicles-companies/
⁷ "Baidu Expands Robotaxi Service Globally with Major Dubai Partnership" - https://www.stocktitan.net/news/BIDU/baidu-s-apollo-go-enters-strategic-partnership-with-dubai-rta-to-yq6kbug9se16.html
⁸ "Zoox Expands Autonomous Vehicle Testing to Los Angeles" - https://www.therobotreport.com/zoox-expands-autonomous-vehicle-testing-los-angeles/
⁹ "Tesla's Bold Move: Model Y Robotaxi Service to Hit Austin's Roads by June 2025!" - https://opentools.ai/news/teslas-bold-move-model-y-robotaxi-service-to-hit-austins-roads-by-june-2025
¹⁰ "Aurora, Continental & NVIDIA Partner to Launch Driverless Trucks by 2027" - https://texasborderbusiness.com/aurora-continental-nvidia-partner-to-launch-driverless-trucks-by-2027/
¹¹ "Autonomous trucking developer Kodiak Robotics to go public via SPAC" - https://www.therobotreport.com/kodiak-robotics-autonomous-trucking-developer-goes-public-via-spac/
¹² "Kodiak Robotics and Atlas Energy Solutions Launch First Commercial Driverless Truck Service in Permian Basin" - https://www.ctol.digital/news/kodiak-robotics-atlas-energy-first-commercial-driverless-truck-permian-basin/
¹³ "Autonomous Driving: Daimler Truck delivers latest iteration of autonomous-ready truck platform to Torc" - https://northamerica.daimlertruck.com/news-stories/2025/autonomous-driving-daimler-truck-delivers-latest-iteration-of-autonomous-ready-truck-platform-to-torc
¹⁴ "NVIDIA DRIVE Hyperion Platform Achieves Critical Automotive Safety and Cybersecurity Milestones for AV Development" - https://investor.nvidia.com/news/press-release-details/2025/NVIDIA-DRIVE-Hyperion-Platform-Achieves-Critical-Automotive-Safety-and-Cybersecurity-Milestones-for-AV-Development/default.aspx
¹⁵ "Will Mobileye be the winning supplier of autonomous driving? Or Waymo?" - https://www.reddit.com/r/SelfDrivingCars/comments/1gmjone/will_mobileye_be_the_winning_supplier_of/
¹⁶ "Mobileye | Driver Assist and Autonomous Driving Technologies" -
https://www.mobileye.com/
¹⁷ "BriefCASE: Revving up - Qualcomm's automotive business hits high gears" - https://www.spglobal.com/automotive-insights/en/blogs/2024/11/briefcase-qualcomms-automotive-business-hits-high-gears
¹⁸ "Why Tesla's camera-only approach may be a mistake" - https://www.fastcompany.co.za/design/why-teslas-camera-only-approach-may-be-a-mistake-6e572ced-d725-4c39-ae21-f74c29ef9900
¹⁹ "Tesla's Vision-Only Strategy Hits a Speed Bump as NHTSA Investigates FSD" - https://opentools.ai/news/teslas-vision-only-strategy-hits-a-speed-bump-as-nhtsa-investigates-fsd
²⁰ "Tesla's FSD Struggles: Is Vision-Only Enough?" - https://evhype.com/teslas-fsd-struggles-is-vision-only-enough/
²¹ "Partially autonomous cars forecast to comprise 10% of new vehicle sales by 2030" - https://www.goldmansachs.com/insights/articles/partially-autonomous-cars-forecast-to-comprise-10-percent-of-new-vehicle-sales-by-2030
²² "Self-Driving Truck Market Set to Hit $41.2 Billion by 2035" - https://shipping.einnews.com/amp/pr_news/793081719/self-driving-truck-market-set-to-hit-41-2-billion-by-2035-key-trends-driving-growth-robotics-inc-plusai-inc
²³ "Waymo - Wikipedia" - https://en.wikipedia.org/wiki/Waymo
²⁴ "Autonomous vehicle technology for self-driving cars can cost up to $100,000 per vehicle, but is expected to drop to around $3,000 by 2035" - https://www.itskrs.its.dot.gov/2020-sc00454
²⁵ "2025: The Defining Year for Autonomous Vehicle Adoption" - https://investorplace.com/hypergrowthinvesting/2025/03/autonomous-vehicles-why-2025-will-usher-in-the-self-driving-car/