How smart is in-car AI really?
Honda recently unveiled new information about its 0 Series, its latest venture into electric vehicles. It announced a new cutting-edge operating system called Asimo, designed to adapt and evolve continuously based on the driver’s preferences. According to Honda, it features a "new level of intelligent vehicle technology." At CES, it demonstrated how a driver interacted with a ‘friend’ chatbot quite reminiscent of the AI from Her.
Though I find their dedication to innovation commendable, I’m always a little hesitant when car brands develop their own software for solutions that are already widely available in the market. It makes sense that Honda wants to control the quality of the chatbot software used in its cars, and that it could learn from all the personal driver data gathered there. But traditional car (primarily hardware) brands hardly ever match the fast innovation culture that software and platform companies tend to showcase.
To give an example: I once tried to use the custom voice assistant of my own car to activate my fog lamps. For some reason, the system flipped over the fact that they were only located in the back. It took me a very long time to convince the system to do what I wanted. The UX was horrendous compared to the ease of use and efficiency of systems like Siri or ChatGPT.
This is one of the biggest conundrums of car companies of today, when it comes to apps and platforms:
The trouble here is that many tend to think in terms of their own needs, for instance: “How can WE gather more data and learn”. The question they should be asking is “What is best for the customer”. 9 out of 10 times, that will be to work with existing external solutions - that their customers already know and love - of which the companies innovate a lot faster than they do.
Just think about it, which of these two options will result in the best CX?:
Option 1: People have a separate car chatbot, a fridge chatbot, a smart lighting chatbot and a different one for all the brands in their smart (home) devices...
OR Option 2: they use the same personal chatbot (and later AI agent) – like ChatGPT or Claude – over all their devices so it knows all of their preferences and needs. Imagine someone driving home, receiving a message from their partner “Don’t forget to buy Hoysin sauce for dinner”, being able to ask the fridge “Do we have Hoysin sauce in the fridge” (because their partner never checks if there still is any) and, if not, receiving a reminder from their car, because they just drove right by the shop, already forgetting what they had to do.
My guess is that everyone will choose option 2, right?
I firmly believe in the opportunities presented by voice interfaces, especially in the automotive industry. But I also believe that the future will be all about integrations and not siloed bespoke approaches that are not interconnected.
To give an example: I once tried to use the custom voice assistant of my own car to activate my fog lamps. For some reason, the system flipped over the fact that they were only located in the back. It took me a very long time to convince the system to do what I wanted. The UX was horrendous compared to the ease of use and efficiency of systems like Siri or ChatGPT.
This is one of the biggest conundrums of car companies of today, when it comes to apps and platforms:
- Do they develop it themselves?
- Or do they partner up with experts?
The trouble here is that many tend to think in terms of their own needs, for instance: “How can WE gather more data and learn”. The question they should be asking is “What is best for the customer”. 9 out of 10 times, that will be to work with existing external solutions - that their customers already know and love - of which the companies innovate a lot faster than they do.
Just think about it, which of these two options will result in the best CX?:
Option 1: People have a separate car chatbot, a fridge chatbot, a smart lighting chatbot and a different one for all the brands in their smart (home) devices...
OR Option 2: they use the same personal chatbot (and later AI agent) – like ChatGPT or Claude – over all their devices so it knows all of their preferences and needs. Imagine someone driving home, receiving a message from their partner “Don’t forget to buy Hoysin sauce for dinner”, being able to ask the fridge “Do we have Hoysin sauce in the fridge” (because their partner never checks if there still is any) and, if not, receiving a reminder from their car, because they just drove right by the shop, already forgetting what they had to do.
My guess is that everyone will choose option 2, right?
I firmly believe in the opportunities presented by voice interfaces, especially in the automotive industry. But I also believe that the future will be all about integrations and not siloed bespoke approaches that are not interconnected.
