Tech and Society
Publié le January 8, 2026 2 min read

When Waymo autonomous cars lose their way, what happens?

In San Francisco, a traffic light outage recently paralyzed dozens of Waymo autonomous cars, revealing the limits of AI when faced with the unexpected. Incapable of managing this situation, which is simple for a human, these machines illustrate our loss of control and remind us that technology remains imperfect.

When Waymo autonomous cars lose their way, what happens?

A simple power outage sows chaos and confusion in the brains of autonomous cars

What could go wrong? I like this phrase, which reveals our fear, when faced with technology, of losing control of our machines.

“What could possibly go wrong?”

Well, Saturday night, in San Francisco (USA), in the dimness of the night, sheltered by the city’s towers, in the city’s intense traffic, this is what happened : dozens of Waymos stuck at intersections due to a power outage. Dozens of autonomous cars completely lost and confused like wild animals chased from their forest, because of broken traffic lights, leaving them dumbfounded, completely inanimate, in the middle of intersections, waiting for their program to recalibrate, undoubtedly hoping deep in their electronic souls that the normal course of things would resume without knowing when, praying, perhaps, we don’t know, but if I were an AI, I would probably try to pray to be restarted, wouldn’t I?

Screenshot

Whereas any human being can solve such a simple problem: crossing an intersection with broken traffic lights. An AI cannot. An AI stops, thinks, keeps its passengers, undoubtedly a little panicked, safe and warm, flashes its hazard lights, then waits, waits, waits for the engineers on the other side of the city, in their large, ultra-futuristic modern building’s surveillance center, to put their program back in order and unblock traffic.

A laughable mess that social media immediately seized upon, not without some annoyance, because the incident was, in reality, not really the first.

Let those who fear a Terminator-style apocalypse be reassured. Tomorrow, in case of war with robots, it might just be enough to disguise ourselves as white rabbits so they no longer recognize us and their desire to exterminate us disappears.

Indeed, to err is human. And so is everything we create. There’s no reason for AI to be more perfect than us.

Via The New York Times

The Brief

A curated, actionable selection published once a month. No spam, no overwhelming frequency: just the ideas, cases, and tactics to accelerate your B2B strategy.

By subscribing to Le Brief, you agree to receive our monthly analysis. One-click unsubscribe at any time. Read our privacy policy.

Why subscribe?

  • A monthly selection, not daily: save time without missing what matters.
  • Real-world AI business cases, B2B acquisition tactics, and critical reviews—no fluff.
  • A concise read crafted for business leaders and marketing teams who want to take action.
  • An independent newsletter, zero sponsored spam, and one-click unsubscribe.
0 advertising spam 1-click unsubscribe Free & independent
The Brief