News of the day

Our take

Hi Dotikers!

Picture this: you're driving, it's pouring rain, and with a simple "Hey, find me a nice coffee shop on the way and add it to my route," your car understands, searches, and adjusts your GPS. That's exactly what General Motors is promising for 2026 by integrating Gemini, Google's AI, into its Chevrolet, Cadillac, Buick, and GMC vehicles.

GM isn't alone in this race either. Mercedes is embedding ChatGPT, Tesla is integrating xAI's Grok, and Stellantis is betting on Mistral. In just a few months, our car cabins have become the new battleground for generative AI giants. After our smartphones, computers, and smart speakers, the automobile is the next territory to conquer.

If you've ever tried talking to your current car's voice assistant, you know the frustration: you have to pronounce commands exactly as it expects, it doesn't understand your accent, and the slightest variation makes it glitch. Dave Richardson, GM's VP, sums up the problem nicely: "Current assistants are trained on precise keywords. With large language models, everything changes. They understand context, remember previous conversations, and adapt to the way you speak."

But GM's ambition goes beyond a simple embedded chatbot. The assistant will be truly integrated into the vehicle's systems: access to maintenance data to alert you before problems arise, preheating the cabin before you get in, route suggestions based on your remaining battery range.

The idea? Create an AI specialized in automobiles, trained specifically on each vehicle's characteristics. "We take a base model and train it on the precise specifications of the vehicle," Richardson explains. Kind of like if ChatGPT had read your car's manual 10,000 times.

And while GM is starting with Gemini, the company is also testing other models, potentially including those from OpenAI and Anthropic. The battle is on.

A.

Tweet of the day