Skip to content

Future Tech/consumer

Kids Born After 2020 Will Not Need a Driver's License

Waymo runs 200,000+ weekly trips with no driver. Tesla, Zoox, and Chinese players scale fast. By 2032 the question is not whether kids learn to drive, but where they would need to.

// By 2032 · medium confidence · disruption 10/10

Prediction

// 2032

By 2032, in major US and Chinese cities, the majority of urban miles will be in autonomous vehicles. Most kids born after 2020 in those cities will never get a driver's license.

Confidencemedium
Disruption10/10

What dies

  • standalone gps
  • the physical car key

Who wins

  • Waymo
  • Tesla
  • Zoox

filed: 2026-05-24 · guptadeepak.com

The hook

Waymo crossed 200,000 paid driverless rides per week in 2025. The number was zero in 2020. The S-curve is steeper than every industry analyst projected, and the 'ten years away forever' joke about self-driving stopped being funny around 2023.

Thesis. Self-driving cars are not a car story. They are a backend infrastructure story: V2X mesh, edge compute fabric, certificate management for billions of vehicle messages, and new categories of attack on physical kinetic systems.

The story

The current state

Waymo runs commercial driverless service across SF, LA, Phoenix, Austin, and Atlanta. Mercedes Drive Pilot ships Level 3 in approved jurisdictions. Tesla FSD v13 shipped late 2024 with end-to-end neural-net driving. Baidu Apollo Go has done more than 10 million cumulative robotaxi rides in China.

The inflection point

Between 2018 and 2022 the AV winter set in: every automaker promised L4 within 24 months, almost everyone missed. The 2023 to 2024 stretch flipped the curve. Waymo expanded into four new cities, Tesla shipped end-to-end neural nets, and Mercedes shipped eyes-off L3 in production sedans.

The prediction

By 2032, in major US and Chinese cities, autonomous vehicles handle the majority of urban miles. Kids born after 2020 in those cities grow up summoning rides the way previous generations called Ubers. A driver's license becomes optional, then quaint, then a hobbyist credential.

Who wins, who loses

Winners: Waymo, Tesla, Zoox, Wayve, Baidu Apollo, plus the V2X infrastructure vendors and the edge compute providers that sit underneath. Losers: standalone GPS, the physical car key, driving schools, parking lot operators, and most independent auto insurers.

Timeline and risks

The technology timeline is on track. The risk is regulatory. One high-profile fleet incident can pause an entire city's service the way Cruise was paused in October 2023. The infrastructure layer (C-V2X, certificate authorities for V2V messages) is also behind the deployment curve.

First signals (verify today)

Waymo at 200,000+ weekly rides across 5+ cities. Tesla FSD v13 shipping. Wayve raised $1B. Mercedes Drive Pilot Level 3 in production.

Key data points

  • Waymo weekly trips: 200,000+ in 2025
  • Mercedes Drive Pilot L3 approved: Nevada, California, parts of EU, 2023 to 2024
  • Tesla FSD v13 shipped: late 2024
  • Wayve raised $1B: 2024 (SoftBank, Nvidia, Microsoft)
  • Baidu Apollo Go cumulative rides: 10M+

Contrarian angle

Most coverage frames self-driving as a car-industry story. The real story is the cybersecurity threat model when a city has 500,000 networked, kinetic, GPS-dependent, certificate-authenticated robots driving among humans. A nation-state attack on a V2X certificate authority could brick urban transport, and we have not built the defensive infrastructure for it.

The flip side

What this kills

The paired obituary in Tech Graveyard.

Read the obituary

FAQ

What is V2X and why does it matter for self-driving?

V2X (vehicle-to-everything) is the cellular and short-range mesh that lets vehicles talk to each other, to roadside infrastructure, and to pedestrians' phones. It matters because perception alone has limits in occluded scenarios; V2X gives cars a non-line-of-sight view of the road.

Why is Waymo scaling faster than Tesla on driverless rides?

Waymo runs a constrained operational design domain (mapped cities, geofenced) with redundant sensors. Tesla runs camera-only on any road, which is a harder problem. Waymo is winning the deployment race; Tesla is making a longer-tail bet on generalization.

Can a robotaxi be hacked while in motion?

The realistic threat is not a single car but the fleet management layer: dispatch systems, certificate issuance for V2X, and OTA update channels. Compromise of any of those is a city-scale incident. Per-vehicle compromise is much harder due to defense-in-depth.

What happens to driving as a hobby?

It survives as a recreational and rural activity. Track days, classic cars, and rural pickup trucks remain. What changes is the default in urban centers, where owning and operating a car becomes optional.

How do self-driving cars handle GPS spoofing or jamming?

Modern AVs fuse GPS with inertial measurement, wheel odometry, visual SLAM, and HD map matching. GPS spoofing is detectable through cross-sensor disagreement. The harder attack vector is the V2X certificate authority, which is the trust root for vehicle-to-vehicle messages.

More from guptadeepak.com

Want the technical deep-dive behind this prediction?

Read the companion article

More from the consumer desk.