Why Waymo’s New Gemini Integration Feels More Like a Gimmick Than a Breakthrough

Waymo's Gemini AI in robotaxis is a gimmick, not a breakthrough. Redundant with phones, it distracts from essential autonomous driving advancements.

4 min. read
Why Waymo’s New Gemini Integration Feels More Like a Gimmick Than a Breakthrough

Why Waymo’s New Gemini Integration Feels More Like a Gimmick Than a Breakthrough

Waymo recently rolled out a passenger-facing software update for its custom-built Ojai robotaxis, adding Google’s Gemini assistant to the cabin screens. While the update brings conversational AI to riders, I see this new feature as an unnecessary gimmick rather than a true leap forward for autonomous driving.

Ever since I was an 11-year-old kid watching Steven Spielberg’s movie Minority Report, I have wanted fully self-driving cars to exist. Back then, I did not even have a driver's license, and my feet could not even reach the gas or brake pedals! Still, I was already tired of sitting in traffic jams and going on long, boring car rides. I always looked forward to the day when I could just get into a car, say where I wanted to go, and lie back to watch a movie.

Today, that future is getting closer. Waymo already has a fleet of over 3,000 self-driving cars, serving around 500,000 paid rides every week across 11 cities. But their latest update for their custom Ojai vehicles—cars built from scratch just for passengers—makes me wonder if we are focusing on the wrong things.

Why do we need another AI assistant in our passenger seats?

Waymo's new update puts a special version of Google’s Gemini assistant onto the screens inside the taxi. Honestly, this feels like a feature we just do not need.

Most of us already carry a great AI assistant right in our pockets. The Google Gemini app on my phone works wonderfully and gets things right most of the time. So why do we need another version of Gemini built into the taxi itself? Many car companies have tried to build their own voice assistants over the years, and none of them seem to work very well. And you know what? That is perfectly fine! We do not need them to.

Do we really need an extra AI assistant in the car just to turn down the air conditioning? To me, this feels more like a fancy trick than a useful feature.

How does Waymo keep its driving system safe from the AI chatbot?

Even though I think the feature is unnecessary, the technology behind how they set it up is actually quite clever. What they have built uses two completely separate computer systems to make sure the passenger assistant can never interfere with the actual driving.

Here is how they split the systems:

  • The Driving Brain: This is the core system that handles the driving. It takes info from cameras, radar, and lasers to track everything on the road. It matches this data with 3D maps to find the car's position and uses smart models to predict how other drivers will behave. Waymo has tested this driving brain over millions of real-world miles.
  • The Passenger Space: This is the system for the screens and the Gemini voice assistant. To protect your privacy, the microphone stays off and does not record anything until you tap the Gemini icon on the screen. Because this system is completely locked away from the driving brain, it cannot see live driving data or send commands to the steering wheel, brakes, or speed controls.

To keep the chatbot on track, they also wrote a massive list of 1,200 safety rules in their code. For example, the AI cannot pretend to be the driver. If you ask how the car sees the road, it must say "The Waymo Driver uses sensors," instead of "I use sensors." It is also limited to short, simple answers. It can change the temperature, lights, and music, but it cannot open the windows, move the seats, change your destination, or buy things for you.

What is the future of the autonomous passenger experience?

Inside the Ojai vehicle, there are three different screens. The new update is smart enough to change how these screens look depending on where you sit. If you are sitting alone in the back, the screen right in front of you shows all the music and temperature controls, while the other two screens go quiet. You can even turn on a "Calm Mode" to dim all the screens and show only a simple map.

This is all highly advanced engineering, but it brings me back to my main point. I say a loud, happy "YES" to any AI system that lets me take my hands off the steering wheel forever. But as for a voice in the car telling me where to find the cheapest slice of pizza? I say, "meh." Let's focus on the driving, and let us use our own phones for the rest.