General Motors is putting ChatGPT into its vehicles

"ChatGPT is going to be in everything.”

American automaker General Motors (GM) plans to integrate a ChatGPT-based voice assistant into its vehicles, giving drivers access to the hugely popular AI chatbot while on the road.

“ChatGPT is going to be in everything,” Scott Miller, GM’s VP of software defined vehicle and operating system, told Reuters.

The status quo: Thanks to systems like Apple CarPlay and Android Auto, it’s easy to connect our phones to our cars. Once linked, we can ask Siri for driving directions or instruct Google Assistant to play a song on Spotify, all while keeping our hands safely on the steering wheel.

“Customers can expect their future vehicles to be far more capable.”

General Motors

A handful of carmakers have even built custom AIs into their vehicles — Tesla drivers can tell their cars to turn on the seat warmers, for example, and BMW owners can ask their voice assistant for an update on the car’s tire pressure.

If a driver needs something beyond these basic commands, though, today’s AIs aren’t going to be able to offer much assistance — and GM wants to change that. 

What’s new? On March 10, Semafor broke the news that GM is developing a new voice assistant for its vehicles, and according to the site’s sources, it’s based on the same AI models powering OpenAI’s ChatGPT, but with a “car-specific” layer added on top.

Miller confirmed this news and added potential uses for the AI: a driver might ask the voice assistant if a lit diagnostic light is serious enough to warrant pulling over, or they might request help changing a tire, which could prompt the AI to play a “how-to” video on the car’s display.

“This shift is not just about one single capability like the evolution of voice commands, but instead means that customers can expect their future vehicles to be far more capable and fresh overall when it comes to emerging technologies,” a GM spokesperson told Reuters.

The new voice assistant is based on the same AI models powering OpenAI’s ChatGPT.

Looking ahead: The collaboration between GM and OpenAI isn’t coming out of nowhere — in 2021, the automaker announced a long-term partnership with Microsoft to commercialize self-driving vehicles, and Microsoft happens to be OpenAI’s biggest investor.

Microsoft, meanwhile, has already integrated the most powerful version of ChatGPT — built on OpenAI’s newly announced GPT-4 model — into its Bing search engine.

GM hasn’t said when it anticipates releasing its new voice assistant or whether it will be built on GPT-4, too, but one of the major advances with that model is that it’s less likely than previous ones to “hallucinate” — that is, write things that sound factually correct but aren’t.

GPT-4 does still hallucinate sometimes, though, and a lot could go wrong if a car’s voice assistant were to confidently misinform a driver on the road — using Miller’s example, they might get into an accident if told it’s safe to continue driving when they should really pull over — so the automaker would be wise to limit the AI to less critical features or wait until it’s sure whatever model it’s using sticks to the facts.

We’d love to hear from you! If you have a comment about this article or if you have a tip for a future Freethink story, please email us at [email protected].

“Treasure map” guides scientists to massive meteorite
A “treasure map” highlighting places where meteorites are most likely to be found has led to the discovery of a 17-pound space rock.
Large language models are biased. Can logic help save them?
MIT researchers trained logic-aware language models to reduce harmful stereotypes like gender and racial biases.
New voice cloning AI lets “you” speak multiple languages
Voice cloning AIs are gaining more abilities, while the amount of audio needed to replicate a person’s voice is shrinking.
New AI can “reimagine” your pictures in infinite ways
Stability AI has launched Stable Diffusion Reimagine, an AI that uses uploaded images as inspiration for new creations.
A search engine researcher explains the promise and peril of letting AI search the web for you
AI responds to questions but doesn’t actually know anything and is prone to making things up. Should you trust is with your internet search?
Up Next
a silhouette of a person on a couch using a laptop
Subscribe to Freethink for more great stories