You Don’t Know Agentic AI Yet, But Qualcomm and Google Want It in Your Car

Global chip-maker Qualcomm announced today that it’s partnering with Google Cloud to help automakers more easily integrate advanced Agentic AI into new vehicles and platforms by integrating Google Gemini models onto the Qualcomm Snapdragon digital chassis. Now, if reading just this far has you wondering, what are Agentic AI, Google Cloud, and Google Gemini, and how can this possibly matter to me? We beg you to let us unpack it all and explain.

Google Cloud

You’ve probably heard of the biggest name in cloud-based data storage, networking, and computing services: Amazon Web Services. AWS currently commands about a third of the market. The other big players are Microsoft Azure and Google Cloud, with the latter being the scrappy newcomer (founded in 2011, five years after AWS and a year after Azure). Azure is popular with big corporate enterprises deeply invested in Microsoft software. Google Cloud ranks strongest in artificial intelligence (AI), machine learning (ML), and analytics, and is emerging as the front-runner for connecting vehicles to the cloud.

See All 3 Photos3

Google Gemini

Gemini refers to a family of AI models designed to understand and process information from disparate inputs, like speech, text, photos, video, audio. A great example of its power: Android phone users can ask Gemini to listen to and identify a song that's either playing from a recording or being sung or hummed live. It converts the sound to a numerical sequence and compares this sequence against the sequences for zillions of songs stored in the cloud to find your answer. (Shazam compares spectrographic representations of actual stored song recordings, so humming or live singing to it doesn’t work.)

Agentic AI

This is artificial intelligence that’s capable of reasoning, planning, and helping accomplish multi-part tasks by analyzing them and breaking them down into smaller steps. It uses various tools and application programming interfaces (APIs) that allow it to connect with different platforms and software, that might, for example, allow it to book appointments or travel. Agentic AI is used in healthcare and cybersecurity, for example, to generate software code, etc.

Arriving in Cars Now

Mercedes-Benz has announced that a Google Cloud-based automotive AI agent built on Google Gemini will appear in its forthcoming CLA-Class. It’s said to be capable of not only recommending a restaurant, but also answering follow-up questions like “does it have good reviews, and what’s the most popular dish there?” (The difference in the CLA system and the new Qualcomm/Google collaboration is that the latter can handle many Google Gemini computations "on the edge," which means onboard the vehicle, in the absence of a cloud connection.) Likewise, Volvo—an early adopter of the Android operating system—is also introducing Google Gemini-powered “conversational AI” in its EX90 EV. These companies have integrated these programs on their own; the Qualcomm Snapdragon/Google Cloud collaboration aims to make such integrations easier going forward.

What Agentic AI Promises to Do for You

With complete access to a vehicle’s onboard data, an AI agent could much more accurately anticipate future system failures, recommend use-based maintenance intervals for critical systems, and diagnose the issue behind a check-engine light, assessing severity, and booking an appointment to fix it. Vacation navigation routing could be asked to incorporate thematic lists of attractions, while also programming EV-charging or fuel stops. An AI agent with vehicle-to-everything (V2X) communications at its disposal could coordinate routing and speed recommendations to avoid heavy traffic, construction, or emergency vehicles, and suggest speed recommendations so you only hit green lights. Such onboard assistants could monitor and manage subscriptions and updates to onboard apps, fully integrate with your home automation—the list goes on and on. And perhaps this new collaboration will help speed its rollout.