On Monday, Waymo announced that anyone in Dallas can now download its app and hail a fully autonomous ride — no invite code, no waitlist, no velvet rope. The blog post was chipper. The tech press ran the headline. Nearly 150,000 riders had cycled through the interest list since the city’s commercial debut in February, the company said, and now the gates were open.

It sounds like a milestone. And in the narrow sense, it is. But 150,000 riders over six months in a metropolitan area of roughly 7.8 million people is not a revolution. It is a rounding error — about 2% of the population, and that counts repeat riders. The median Dallas resident has never seen a Waymo, let alone ridden in one. The median Dallas resident probably couldn’t tell you where the service area begins or ends.

That gap — between the press-release future and the on-the-ground present — is the real story of autonomous vehicles in 2026. And it is not a story the industry is eager to tell.

The 2% Milestone

One hundred fifty thousand is a real number, and it is not nothing. It suggests a core of curious early adopters, tech workers, and downtown residents willing to trust a robot at the wheel. But put it next to the 7.8 million people who live in Dallas-Fort Worth, and the scale of the thing snaps into focus. Waymo has been operating in Phoenix since 2020 — six years — and the service area there still doesn’t cover the whole metro. San Francisco went fully public in mid-2024, and you can still find San Franciscans who have never taken a ride.

This is not a criticism of the technology. The cars work. They are, by most accounts, cautious to a fault and impressively safe on surface streets. The criticism is of the narrative that treats each city opening as a step function — as though flipping a switch from “waitlist” to “open” transforms a city’s relationship with transportation. It doesn’t. It transforms the relationship for the sliver of people who live, work, and play inside a geofenced polygon that, in Dallas, still excludes freeways and the airport.

“People think we flipped a switch and the whole city is served,” said a maintenance supervisor at the Avis depot that handles Waymo’s Dallas fleet, speaking on condition his employer not be named. “We’ve got a couple hundred cars cleaning and charging every night. They’re running a loop that covers maybe 15% of the metro. The rest of Dallas might as well be on Mars.”

A City Without Freeways

Here is what “open to all” does not mean in Dallas, at least not yet: It does not mean you can take a Waymo to Dallas Love Field. It does not mean you can hop on I-35 and let the car handle 70-mile-per-hour traffic. Waymo’s own announcement notes that freeway testing and airport service are “queued as the next steps” — which is corporate-speak for “we’re working on it, and we don’t know when it’ll be ready.”

This matters because Dallas is not San Francisco. It is not a compact, 49-square-mile peninsula where surface streets can plausibly get you most places you want to go. Dallas is a sprawling, highway-dependent metro where the airport run is a basic utility, not a luxury. A robotaxi service that can’t use freeways and can’t reach the airport is a novelty, not infrastructure. It is a very impressive demo that happens to charge fares.

None of this is Waymo’s fault, exactly. Freeway driving at speed introduces a different order of risk — longer stopping distances, higher consequences for sensor error, merging scenarios that remain genuinely hard for machine-learning systems. The company is right to be careful. But the carefulness is the point. The carefulness is what the “AVs are here” story keeps eliding.

The Long Tail of the Revolution

Waymo now operates in 10 cities and has said it aims to be in more than 20 by the end of 2026. That is real growth. But each new city requires months of mapping, testing, and phased rollouts before a single paying rider gets in. The fleet in each city numbers in the hundreds, not the thousands. The capital costs — those Jaguar I-Paces, the lidar stacks, the compute, the depot operations outsourced to Avis — are enormous and largely invisible in Alphabet’s consolidated financials.

The business model question is not whether the technology works. It is whether the unit economics work at a scale that actually changes how a city moves. And the answer, for now, is that nobody knows — because nobody has done it at that scale. Not in Phoenix, not in San Francisco, and not in Dallas.

What Monday’s announcement really marked was the end of the beginning. The waitlist is gone. The real test — converting the other 98% of Dallas — hasn’t even started.

Sources