California granted Waymo permission to operate across 47,493 square miles—but the company runs just 2,500 vehicles in five cities. The gap between regulatory approval and actual deployment reveals unit economics Waymo won't disclose and local opposition the state ignores.
Google kept AI Mode ad-free for a year to compete with ChatGPT. Last week, ads appeared. The retreat reveals more than monetization strategy—it shows Google's $175B advertising business makes competing structurally impossible.
Google must double AI compute every six months to meet demand, but capacity limits are already capping product rollouts and revenue. As bubble concerns mount, the company bets that underinvesting poses the greater risk in the infrastructure arms race.
California granted Waymo permission to operate across 47,493 square miles—but the company runs just 2,500 vehicles in five cities. The gap between regulatory approval and actual deployment reveals unit economics Waymo won't disclose and local opposition the state ignores.
California regulators handed Waymo the keys to an Austria-sized territory last week. The Department of Motor Vehicles approved driverless operations across 47,493 square miles spanning 18 counties, from Sacramento wine country to the San Diego border. Coverage area increased tenfold overnight.
Waymo operates 2,500 vehicles in five cities.
The gap between permission and deployment matters. Regulatory approvals generate headlines about autonomous transformation. Actual operations expand block by block, limited by vehicle production, charging infrastructure, and economics Waymo won't disclose. The DMV granted access to 40 million Californians. Waymo currently serves paying riders in parts of San Francisco, parts of Los Angeles, parts of Phoenix, Atlanta, and Austin.
"We're looking forward to opening our service to more Californians," per Waymo's November 21 statement. No launch timeline. San Diego gets "mid-2026" as a target. Sacramento, Wine Country, the North Bay, everything between LA and San Diego? Nothing specific.
This matters because the gap between regulatory theater and operational reality shapes how quickly autonomous ride-hailing actually scales. Also because California's state-level approval structure explicitly prevents cities from evaluating local impacts, blocking municipal governments even as they watch taxi drivers, Uber operators, and transit workers face displacement.
The Breakdown
• California DMV approved Waymo for 47,493 square miles across 18 counties, but only 2,500 vehicles operate in five cities
• Operating costs estimated at $2/mile while charging $1.60-$2.60/mile leaves minimal margin before R&D and infrastructure costs
• State regulatory structure blocks municipalities from oversight, rendering San Diego's unanimous opposition legally meaningless
• Fleet utilization of 13 rides per vehicle daily falls far below human-driven ride-hail density in same markets
Waymo now holds permits covering territory larger than Austria or South Carolina. Actual service? Maybe 200 square miles of operational footprint, being generous. One million rides monthly sounds impressive. Do the math: 2,500 vehicles, 30 days, that's about 13 rides per car daily.
Uber floods San Francisco with thousands of cars during peak hours. Each one cycling through dozens of rides. Waymo announced geographic expansion. What about fleet expansion? Silence on that.
The DMV approval includes all nine Bay Area counties, five Southern California counties, scattered territories north to Cloverdale and east to Sacramento. Waymo can test driverless vehicles across this footprint. Offering paid rides requires separate California Public Utilities Commission approval for each service area. Most newly-approved territories? No CPUC applications filed.
Wine Country is different from urban cores. Napa and Sonoma spread across 2,500 square miles. Remote vineyards. Winding roads. Uber and Lyft barely work there now. Waymo would need charging infrastructure, vehicle cleaning facilities, support operations for dispersed rural destinations. Capital investment with no clear payback timeline.
"Where there are opportunities are getting people between the different attractions, especially wineries, and getting what we call 'up valley' into Glen Ellen, Kenwood and the springs," Tim Zahner from the Sonoma Valley Visitor Bureau told local media. That optimism assumes Waymo deploys serious fleet capacity where ride density will run a fraction of urban markets.
Alphabet has deep pockets. But the operational reality of autonomous ride-hail requires physical infrastructure in every service territory: vehicle staging areas, charging stations, cleaning and maintenance facilities, remote assistance centers staffed 24/7. Regulatory approval costs nothing. Deployment infrastructure costs millions per market.
Uber put Waymo's operating costs at roughly $2 per mile in analyst presentations. Waymo's pricing runs $1.60 to $2.60 per mile by location and demand. Flag drop fees add about $10. Even at the high end, margins look thin before R&D amortization, fleet depreciation, infrastructure buildout across 18 counties.
One Reddit analysis pegged operating costs at $60,000 per vehicle annually. About $164 daily. If 13 rides generate $200 to $300 gross revenue, where's the profit?
Waymo has never disclosed profitability metrics. Alphabet doesn't break out Waymo financials in earnings reports, bundling autonomous vehicle operations into "Other Bets" that collectively lost $3.3 billion in 2024. The company occasionally releases safety statistics and monthly ride volumes. Revenue, profit margins, and unit economics remain opaque.
This opacity matters because Waymo positions itself as offering robotaxi service at scale. One million monthly rides across five cities demonstrates technical capability. Sustainable economics would demonstrate business viability. The distinction determines whether autonomous ride-hailing becomes a real transportation mode or an expensive technology demonstration subsidized indefinitely by Alphabet.
Consider the infrastructure investments required for the newly-approved territories. Each service area needs vehicle depots with charging capacity for dozens or hundreds of vehicles. Waymo operates all-electric fleets of Jaguar I-Pace and Chinese-made Zeekr RT vehicles. Charging infrastructure alone runs into millions per location. Add vehicle cleaning, maintenance bays, remote operations centers, and local staff for fleet management.
San Diego's Metropolitan Transit System Taxi Advisory Committee voted last month to oppose Waymo's planned 2026 launch. The committee called for prohibiting driverless vehicles at San Diego International Airport and urged state legislators to restore local control over autonomous vehicle deployment. That opposition will not stop Waymo. California's regulatory structure gives cities no authority over AV operations.
But opposition signals that deployment beyond tech-friendly strongholds faces resistance Waymo hasn't encountered in San Francisco. Labor groups in San Diego, taxi cooperatives, and food delivery workers all see displacement coming. They can't stop state-approved operations. They can make market entry politically messier and operationally more complex.
The economics of serving San Diego differ from serving San Francisco. Lower population density, more spread-out destinations, less late-night demand, fewer dense commercial corridors where vehicles can chain multiple short trips. Utilization rates will almost certainly run lower. Operating costs per mile won't decrease. The unit economics problem gets harder, not easier, as Waymo expands into secondary markets.
California explicitly prevents municipalities from regulating autonomous vehicles. The DMV controls operational safety permits. The CPUC controls commercial service authorization. Cities are preempted from enacting local ordinances that encroach on state authority over AVs.
This was intentional. State legislators saw fragmented municipal control as innovation barrier, recipe for what one industry observer called "highly fragmented municipalities" creating "a lot of free-rider vetos." The state grabbed authority.
San Diego City Councilmember Sean Elo-Rivera chairs the Metropolitan Transit System's Taxi Advisory Committee. Last month the committee voted unanimously against Waymo's expansion. They urged state legislature and Governor Newsom to restore local control over autonomous vehicles in their jurisdictions. Legally meaningless.
"My chief concern is what this means for people who are keeping a roof over their head and feeding their families by being drivers," Elo-Rivera told the San Diego Union-Tribune. The committee wanted formal protests to CPUC and DMV. They wanted driverless vehicles banned at San Diego airport. They wanted communities to vote on AV deployment.
None of it matters. The committee lacks authority. San Diego and Imperial Counties Labor Council opposes robotaxis. United Taxi Workers opposes robotaxis. Mayor Todd Gloria supports Waymo. State regulators will approve operations.
SB-915 tried to change this in 2022. The bill would have allowed cities to permit or regulate autonomous vehicles. Failed in the California legislature. Cities couldn't regulate AVs before. They still can't.
That regulatory framework benefits companies like Waymo enormously. Rather than negotiating with dozens of municipal governments, winning approval from local transit authorities, and addressing community concerns city by city, Waymo secures two state permits and operates anywhere those permits allow. The DMV grants operational authority. The CPUC grants commercial authority. Cities watch.
The efficiency comes with tradeoffs. Local governments cannot require AV operators to demonstrate they won't displace workers, won't create safety hazards in specific neighborhoods, won't flood airports with vehicles undermining existing ground transportation systems. Municipal oversight vanished by design.
San Diego's opposition centered on exactly those concerns. Taxi drivers fear income loss. Airport workers want protection for existing jobs. The city has no mechanism to require Waymo to address those impacts before launching service. State-level regulation optimizes for industry expansion. Local regulation would have optimized for community control.
For municipalities and workers: California's regulatory structure treats autonomous vehicle deployment as purely a technical question to be answered by state agencies. Labor displacement, local traffic patterns, and community preferences don't factor into approval decisions. Cities that want to slow deployment, study impacts, or require operator commitments have no authority. Workers facing job loss have no seat at the table. The state chose innovation velocity over local input.
For Waymo and competitors: Geographic approvals are cheap. Operational deployment is expensive. Waymo now has permission to operate across 47,493 square miles but actual expansion requires fleet growth, infrastructure buildout, and solving unit economics that remain undisclosed. The company generates impressive ride volumes with a relatively small fleet, but scale demands either radically lower operating costs or sustained willingness to run money-losing operations for years. Every new market adds fixed infrastructure costs while utilization rates likely decrease outside dense urban cores. The math gets harder as territories expand.
For understanding AV economics: The industry spent a decade positioning autonomous ride-hailing as inevitable. Waymo is closest to proving the technology works commercially. But technical function differs from economic viability. The company won't disclose profitability while announcing aggressive expansion. That suggests unfinished business model. Without transparent unit economics, the core questions remain unanswered. Will robotaxis undercut human drivers on price? Do they need permanent subsidies? Does massive scale finally deliver profit? Regulatory approvals sidestep these questions entirely.
Q: Why can't San Diego block Waymo if their transit committee voted against it?
A: California law gives exclusive authority over autonomous vehicles to state agencies—the DMV controls safety permits and the CPUC controls commercial operations. Cities are legally preempted from regulating AVs. When SB-915 tried to restore local control in 2022, it failed in the legislature. San Diego's unanimous opposition has no legal effect.
Q: How much does a Waymo ride cost compared to Uber or Lyft?
A: Waymo charges $1.60 to $2.60 per mile plus a $10 flag drop fee, typically running about 50% more expensive than Lyft in San Francisco. Users report paying the premium for reliability—Waymo vehicles don't cancel rides at late hours or from unpopular pickup locations, a common complaint with human-driven services.
Q: What's the difference between DMV approval and actually offering paid rides?
A: The DMV permit allows driverless testing across 47,493 square miles. Operating a commercial robotaxi service requires a separate California Public Utilities Commission permit for each specific area. Waymo hasn't filed CPUC applications for most newly-approved territories, meaning DMV approval doesn't equal imminent service launch.
Q: Why is expanding to Wine Country harder than operating in cities?
A: Napa and Sonoma counties cover 2,500 square miles of remote vineyards with low population density. Each service area requires charging infrastructure, vehicle cleaning facilities, and 24/7 support operations. With far fewer potential rides per square mile than urban markets, the infrastructure investment faces unclear payback timelines.
Q: How many Waymo vehicles are actually on the road right now?
A: Waymo operates 2,500 vehicles across five cities: San Francisco, Los Angeles, Phoenix, Atlanta, and Austin. These vehicles complete about 1 million rides monthly, averaging roughly 13 rides per vehicle per day. For comparison, Uber saturates major cities with thousands of human-driven vehicles that each handle dozens of daily rides.
Marcus Schuler
Robert Brown
Google kept AI Mode ad-free for a year to compete with ChatGPT. Last week, ads appeared. The retreat reveals more than monetization strategy—it shows Google's $175B advertising business makes competing structurally impossible.
Google must double AI compute every six months to meet demand, but capacity limits are already capping product rollouts and revenue. As bubble concerns mount, the company bets that underinvesting poses the greater risk in the infrastructure arms race.
Musk's Grok chatbot ranked him fitter than LeBron James and smarter than Einstein this week before xAI deleted the posts. The sycophantic episode reveals centralized AI control mechanisms in a system now wired into federal intelligence workflows.
Google's new image generator embeds "critical" authenticity watermarks in every AI creation. But pay $19.99/month and the visible mark disappears. The company monetizes the transparency tools it says protect public trust, while actual deepfake threats go unaddressed.
Get the 5-minute Silicon Valley AI briefing, every weekday morning — free.
