The Day AI Ceases to Be a 'Feature' | This Week's CHINA CASE Deep Dive
Weekly China Automotive Industry News Commentary
July 27, 2026 – August 02, 2026
This is a report based on a slightly edited version of the video content.
You have had the worst day at work and climb into your car, completely exhausted.
Before you even touch anything, the interior lighting dims slightly, and your favorite relaxing music begins to play. Then, the car suggests, 'Would you like to take a scenic route home today?'
A while ago, this would have sounded like science fiction.
However, in the Chinese automotive market, this world is becoming a reality.
What is happening now is not merely the sophistication of in-vehicle software. AI itself is beginning to enter a stage where it redefines 'what a car is.'
In this CHINA CASE Deep Dive, we interpret the changes progressing in the Chinese automotive industry from the perspective of how AI is changing cars and how consumers are beginning to perceive it.
Consumers are not interested in 'AI'
In China today, new technologies and brands centered on AI and advanced driver assistance, such as HIMA (Harmony Intelligent Mobility Alliance), are appearing one after another.
However, interestingly, consumers are not necessarily strongly interested in the technical names or mechanisms of AI itself.
What is important to the user is
the single point of 'how it will make my life more convenient.'
No matter how advanced the algorithm installed, it does not necessarily become consumer value in itself.
For this reason, manufacturers are also trying to evolve AI from a mere 'convenient feature' into something where the user can feel its very existence.
Giving AI a 'body'
One symbol of this is the in-vehicle robot 'Eva' from Lynk & Co.
Instead of treating a Large Language Model (LLM) as a program existing deep within the screen, by giving it facial expressions and physical representation, humans can communicate with AI more intuitively.
Furthermore, initiatives like 'AIVA' are introducing the concept of 'AI-native cars,' where the vehicle itself is designed around AI from the start, rather than simply adding AI to existing automobiles.
This is different in meaning from simply enhancing the performance of a smartphone's voice assistant.
It is the idea that AI is not just a feature of the car, but that the car itself becomes the body of the AI instead.
'Understand, but do not monitor'
However, if cars come to understand human emotions, new problems will also arise.
Being tired. Being in a bad mood. Feeling stressed. Looking sleepy.
A world where a car constantly grasps such states is convenient, but it also carries the risk of making people feel like they are being 'monitored'.
This is where the sense of distance of 'understanding, but not monitoring' becomes important.
In the philosophy AIVA aims for, rather than continuously sending all data regarding user emotions to the cloud, it is processed within the car's local environment as much as possible.
While ensuring privacy, it understands the human state and changes the HMI (Human Machine Interface) accordingly.
In the era of AI automobiles, not only 'how much it can understand humans,' but also how much it is permitted to understand will become part of the product's value.
Understanding the physical world as aimed for by WeRide's 'WIT'
While AI that understands humans inside the car is evolving, AI that understands the world outside the car is also changing significantly.
'WIT,' developed by WeRide, is one such example.
What is important is that it does not simply recognize what is reflected in the camera as an 'object'.
For example, suppose a ball rolls onto the road.
A conventional system would recognize it as a single obstacle.
However, a human would think,
'A child might chase after the ball and jump out after this.'
That is how they think.
In other words, it infers not only the objects present before it, but also the causal relationships of the real world behind them.
Helping AI understand the context of the physical world is becoming crucial for the next generation of autonomous driving.
Unifying 'Perception, Judgment, and Operation'
To accelerate this complex processing, an approach called 'one-stage end-to-end' has emerged.
In conventional autonomous driving, there were multiple steps: perceiving with sensors, judging based on rules or algorithms, planning the route, and operating the steering and brakes.
With end-to-end, these are connected by a single AI model, attempting to link input information directly to driving operations.
For humans, this is closer to the sensation of moving reflexively rather than looking at something and searching through a rulebook in one's head step by step.
Such technology is currently being tested globally in countries like Germany, France, and Japan.
They are attempting to build not only a 'brain that understands the world' for AI, but also 'reflexes' to instantly translate that into action.
Excellent AI Alone Cannot Realize Robotaxis
However, no matter how advanced an AI is created, it alone will not complete an autonomous driving society.
The initiative of the Robotaxi 'HR-1' deployed by Dongfeng Nissan and Hello demonstrates the challenges that lie beyond that.
To establish Robotaxis as social infrastructure, the vehicle itself is not enough.
A massive system is required, including mass production, vehicle management, operations, maintenance, and service management.
It is the same as how developing the best cooking recipe does not mean you can operate a huge restaurant chain.
The 'brain' that is AI. The 'body' that is the automobile. And the 'infrastructure' that keeps them running continuously in society.
Only when these three are in place can AI mobility proceed to social implementation.
'Zombie EVs' Increasing Behind the AI Competition
And here lies another reality currently facing the Chinese automotive market.
Cutting-edge AI development requires enormous research and development costs, computational resources, driving data, and software talent.
Not every automaker can participate in this competition.
As a result, intense polarization is progressing within the Chinese automotive industry.
On one side, there are companies building massive AI-centric ecosystems and attempting to deploy their technology to the global market.
On the other side, there are companies that cannot secure sufficient sales within China and are seeking a way out overseas to survive.
Zotye and Skyworth are cited as symbols of this.
Some of these companies are even moving into emerging markets like India by utilizing a knockdown method, where they export vehicles as parts to be assembled locally rather than exporting finished cars.
This is to survive in new markets while avoiding the high tariffs imposed on finished vehicles.
It is a completely different way of fighting compared to the cutting-edge AI competition.
Intense polarization in the Chinese EV market
When you look at the big picture like this, you can see what is happening in the current Chinese automotive market.
On one hand, cars are being born that use AI to understand human emotions, interpret causal relationships in the physical world, and drive themselves.
Massive AI ecosystems and Robotaxi operation networks for that purpose have even begun to be built.
On the other hand, manufacturers that can no longer survive in the domestic market are seeking a way out in emerging countries, even if it means breaking down vehicles into parts.
AI-native cars attempting to lead the world, and 'zombie EVs' searching for mere survival.
The simultaneous existence of these two extremes is exactly what symbolizes the current Chinese automotive industry.
When we stop being conscious of AI, the real change begins
And there is one more important change.
AI is gradually passing the stage where products sell just by promoting that they are 'AI-equipped'.
What consumers are looking for is not AI itself. It is having a comfortable environment created when they are tired. It is driving safely by predicting danger. It is being taken to their destination reliably without having to drive themselves.
Consumers may eventually stop being conscious of whether that is being realized by AI or by some other technology.
Just as we do not think about the mechanics of breathing when we inhale, the moment we stop being conscious of AI altogether may be the moment AI truly permeates the automobile in the truest sense.
At that point, another question arises.
If a car were to become like a 'best friend' that understands your daily fatigue and even your mood, anticipating what you want before you even realize it yourself, would humans choose that convenience even at the cost of giving up the freedom to drive?
Or would we feel a sense of fear toward a machine that knows far too much about us?
The rapid AI automobile experiments underway in China are beginning to question not only the future of technology, but the very relationship between humans and automobiles.
#China
#Automobile
#Mobility
#ChinaMobility
#TraditionalChineseMedicineBodyView
#ChinaMobilityAndTCMBodyView
#ElectricVehicle
#EV
#BEV
#REEV
#EREV
#RangeExtenderEV
#PlugInHybrid
#PHEV
#Hybrid
#HEV
#ICEV
#AutonomousDriving
#HandsFreeDriving
#SmartDriving
#EndToEnd
#E2E
#AIDriving
#SmartCockpit
#SmartLounge
#AISpeaker
#AIAssistant
#AIAgent
#Case
#SmartCar
#SDV
#AIDV
#AI
#EmbodiedAI
#PhysicalAI
#DailyNote
#DailyUpdate
#DailyPost
#NoteDailyUpdate
