The resentment faced by Baidu's Apollo Go robotaxi: unjust or deserved?
As Baidu's stock soars on Robotaxi demand, drivers fear job loss
All the spotlights have been on Baidu's newly launched robotaxi, Apollo Go, in the past few days. Orders for Apollo Go in Wuhan, one of the first cities to pilot commercial robotaxi services in China, have surged. As of June, Apollo Go has over 300 robotaxis operating in Wuhan from 6 am to 2 am the next day, and Baidu plans to increase the number to over 1,000 by the end of 2024.
Baidu's stock surged on July 9 and continued to gain in the following days supported by the boom in robotaxi orders. Analysts from Guotai Junan Securities expect that the robotaxi service will break even in Wuhan this year—marking the first promising commercial success and synergy between Baidu's AI model capabilities after many years of investment. Many netizens have given positive feedback and rushed to Wuhan to try out the new robotaxi, commenting, "Finally, I don't need to listen to the driver's chitchat," "The robotaxi won't detour to charge me more money," and "The robotaxi is so much cheaper," among other remarks.
Meanwhile, the commercial buzz raises concerns about job security among drivers, with many petitioning Wuhan's transport authority to limit the use of Apollo Go in the city of over 13 million. A recent accident involving Apollo Go hitting a jaywalking pedestrian also raised concerns about whether robotaxis can ensure safety and accommodate unexpected traffic conditions in the megacity.
Job security remains the most controversial topic surrounding the robotaxi, especially under current circumstances. Data from the China Ministry of Transport shows that as of the end of December 2023, the number of registered ride-hailing drivers reached 6.572 million, with 1.482 million new drivers added in 2023 alone. Internet ride-hailing has become a reservoir for jobs for many, and many cities have been warning that the ride-hailing job market has almost reached saturation. The introduction of commercialized robotaxis under these conditions, and at even cheaper prices, has undoubtedly stirred concerns among local drivers already in a highly competitive market.
In today's newsletter, I want to translate an excellent article written by award-winning blogger Zongming Yu that breaks down the nuances behind the robotaxi boom in Wuhan. The article discusses how government policy support is promoting robotaxi innovation, the commercial model, citizen feedback, and the social implications for local drivers and the broader job security debate in the AI era.
IMPORTANT Community Announcement
We are planning to launch a paying member-only online community for you to exchange ideas. We have yet to decide on details, but very likely it will be a Discord community. We welcome other suggestions from you.
Starting from July 15, we will increase our subscription price from $100/year to $188/year, $15/month to $20/month. Prices will NOT increase for paying subscribers who sign up by July 15th, so please subscribe before then to lock in the old price.
Paying members of Baiguan enjoy free access to China Translated, the personal newsletter of Robert Wu, which is currently worth $100/year. Please DM us if you wish for this access.
*BigOne Lab provides consumer sentiment analysis on major social media platforms in China, including Xiaohongshu and Douyin. Institutional clients and investors interested in analyzing sentiments regarding Baidu's robotaxi can contact us at more@bigonelab.com for a quote.
Below is the translation of the original article, "萝卜快跑承受的『冤与怨』," titled "The Unjust Accusations and Resentment Faced by Baidu's Apollo Go Robotaxi."
The unjust accusations and resentment faced by Baidu's Apollo Go robotaxi
Apollo Go robotaxi is hitting the roads in batches, and the timing couldn't be more perfect for China to seize the initiative in the autonomous driving arena. Yet, it's also a challenging time because it might snatch opportunities from ride-hailing drivers — currently, driving for ride-hailing serves as a fallback job for many.
Those who haven't cried deeply in the late night are insufficient to discuss life. Those who haven't experienced the speed and passion of Wuhan buses late at night are insufficient to talk about life in Wuhan.
Speaking of urban legends in Wuhan, alongside Aeroflot planes and Chongqing taxis, Wuhan buses, known as one of the world's three transportation wonders, are undoubtedly an unavoidable presence.
Having lived in Wuhan for six years, I have some insights into this: Years ago, I couldn't understand why Wuhan's Bus 521 was dubbed the "F1 of buses." I thought the stories of bent handrails and flaming tires were just ancient rumors until one time it took me from Simeimen to Huazhong University of Science and Technology in 30 minutes, thoroughly convincing me of its "fighter jet in the sky, Bus 521 on the ground" reputation.
Every driver of Wuhan's Bus 521 has a racer's heart.
But these days, I've been pondering: What would happen if Bus 521 encounters Apollo Go robotaxis, easily seen on the streets of Wuhan Economic Development Zone?
During this period, as Apollo Go robotaxis are deployed in batches on the streets of Wuhan, they have frequently become a hot topic among the public: a must-visit project in Wuhan, skyrocketing orders (with some vehicles peaking at over 20 orders per day), frequent complaints, accused of causing traffic congestion in Wuhan, immediate police reports, alleged manipulation by over ten suspects controlling hundreds of accounts to smear themselves, and reports of collisions with pedestrians running red lights on Wuhan streets... all these have brought it intense scrutiny.
After Bus 521, the Apollo Go robotaxis has become a new topic of discussion in Wuhan's transportation and travel.
The video shows footage taken by a family member in Wuhan while riding in an Apollo Go robotaxi. He said the "Apollo Go" is very slow, but it strictly follows traffic rules.
The question arises: Is Apollo Go robotaxi running the right race?
From current observations, Apollo Go robotaxi's timing is both perfect and imperfect.
To some, it should run even faster — ideally leaving Tesla behind.
To others, its problem is that it runs too fast — will there be any livelihood left for ride-hailing and taxi drivers if this continues?
It's foreseeable that Robin Li (CEO of Baidu), the ultimate "driver" of Apollo Go robotaxi, will inevitably bear more praise and criticism for this.
01
For Wuhan and Baidu, the deployment of Apollo Go robotaxi is undoubtedly timely.
Everyone knows that in April of this year, "Praise China Group" leader Elon Musk once again visited China. [Baiguan: Some Chinese netizens refer to Musk as the "chat group leader" of the "Praise China Group" (中国夸夸群), referencing the many positive comments Musk has publicly made about China.] It is said that his purpose this time was to promote the landing of fully autonomous driving in China.
Autonomous driving is Musk's "nuclear weapon" to prevent Tesla from falling out of the top seven giants in the US stock market and a representative cutting-edge technology bet by many Internet and automobile companies.
Apart from Tesla, the leader in autonomous driving abroad is Waymo, a self-driving travel service provider under Alphabet (Google's parent company).
In the second half of 2023, Waymo and Cruise, a subsidiary of General Motors, obtained licenses to operate paid services in the San Francisco area with their self-driving taxis. However, just three months later, Cruise's self-driving taxis were temporarily suspended due to multiple incidents such as engine failure, collision with fire trucks, blocking ambulances, and dragging pedestrians.
Waymo's valuation also plummeted from a peak of $175 billion in 2018 to about $30 billion by mid-last year.
The concept of autonomous driving in China has been hot and cold. It was hot for a while a few years ago, but it started to cool down in 2019. Around 2022, it warmed up again due to Huawei's entry, but still faced criticism from BYD Chairman Wang Chuanfu — "Autonomous driving is nonsense, those fanciful things are just scams." By 2023, the limelight was on AI generated content, and autonomous driving was nothing.
Due to the low fault tolerance rate in long-tail scenarios (an accident could endanger lives), the difficulty of large-scale landing of autonomous driving is extremely high.
Kevin Kelly once said: (In terms of autonomous driving) We cannot settle for 99% safety, we must pursue 99.99% accuracy. Achieving 98% safety is not difficult, but the final part is the most difficult and time-consuming. It can be said to be extremely difficult to make a breakthrough in the last 2%.
Tesla claims to have the technical capability to achieve Level 3 autonomous driving, but has not yet been fully approved; many domestic car companies are still playing the word game of "L2.999999..." — intelligent assisted driving below Level 3 requires hands-on steering.
Whoever can achieve a large-scale landing of autonomous driving as soon as possible will get a ticket to the future.
On July 3, the day before the opening of the 2024 World Artificial Intelligence Conference, the Ministry of Industry and Information Technology and other five departments clarified Beijing, Shanghai, Shenzhen, Guangzhou, Wuhan and other 20 cities as smart IoT vehicles "vehicle road cloud integration" application pilot cities.
These days, Beijing has publicly solicited opinions on the "Beijing Autonomous Driving Vehicle Regulations (Draft for Solicitation of Comments)", planning to support autonomous driving vehicles for urban public electric vehicle passenger transport, ride-hailing, car rental and other urban travel services.
In short, seizing the heights of next-generation technology cannot afford to be "wise after the event".
02
To Wuhan and Baidu, the effort to promote autonomous driving has a different meaning.
No city in China has supported Apollo Go robotaxi as vigorously as Wuhan — by early this year, Wuhan's smart IoT vehicles had accumulated test road mileage exceeding 3,378.73 kilometers (one-way mileage), covering an area of approximately 3,000 square kilometers, far surpassing the second-place Silicon Valley, equivalent to six San Francisco.
Behind this is Wuhan's strategy to compete for the title of "Global Autonomous Driving First City".
For many years, automobiles have been Wuhan's largest pillar industry. Counting China's automobile industry towns, Wuhan Economic Development Zone, with 14 complete vehicle factories, is an indispensable part.
In 2022, the "武襄十随 Wuxiang Ten Follows" automobile industry cluster led by Wuhan was selected as the list of advanced manufacturing industry clusters by the Ministry of Industry and Information Technology, alongside Shanghai's new energy vehicle cluster and Changchun's automobile cluster.
For many years, automobiles have been Wuhan's largest pillar industry.
But in the face of the new energy vehicle's "new VS old" dualistic narrative, pulling gasoline cars down, and the iterative change from "one engine + four sofas" to "one mobile phone + color TV refrigerator large sofa", Wuhan has been injured—missing out on the "oil-to-electricity" opportunity, Wuhan failed to take the lead in entering the car electrification market.
In March last year, Wuhan led the "price reduction tide" that swept the car market, but this could not change the situation where Hubei's (Wuhan is the provincial capital of Hubei province) automobile production in the first three quarters (1.3098 million vehicles) fell by 10.1% year-on-year, lagging behind Anhui. It is for this reason that the Wuhan Economic Development Zone has announced the slogan "Second Entrepreneurship".
But if electrification is seen as the first half of the competition for new energy vehicles and intelligence is seen as the second half, then Wuhan, who missed the first half, obviously intends to regain its leadership advantage in the second half.
Therefore, Wuhan has the longest autonomous driving test mileage in the world, has become the first city to allow domestic robotaxis to run on highways, and now supports the large-scale deployment of Apollo Go robotaxi's sixth-generation driverless vehicles.
No domestic technology giant needs to rely on Robotaix for its AI big model to produce mature application scenarios as urgently as Baidu IDG (Intelligent Driving Business Group).
Over the years, Baidu IDG (Intelligent Driving Group) has become like the West Point of China's autonomous driving field. However, after several years, the IDG's three main business divisions— the "Autonomous Driving Division" (focusing on Robotaxi) based on the Apollo Go robotaxi, the "Intelligent Vehicle Division" (mainly targeting car manufacturers) focused on mass-produced assisted driving and intelligent cockpits, and the "Intelligent Transportation Division" (To G) primarily providing vehicle-road coordination solutions—are now almost solely dependent on Apollo Go robotaxi as the last remaining "seedling.

As a pioneer in China's "AI + autonomous driving" field, Baidu began its autonomous driving layout in 2013 and launched the world's first autonomous driving open platform, Apollo, in 2017. With the ambition to become the world's largest autonomous driving service provider, Baidu can only place all its hopes on Apollo Go robotaxi.
It can't afford to lose.
Those who hope for China to advance faster in the field of autonomous driving likely wish that Apollo Go robotaxi could run even faster — if not to be far ahead, at least to lead the way.
03
For some ride-hailing and taxi drivers, it seems like Apollo Go Robotaxi hitting the road is coming at quite an inconvenient time.
In recent days, "Apollo Go Robotaxi is taking away drivers' livelihoods" has become a common reason why many oppose Apollo Go Robotaxi hitting the road.
It's hard to say it's all smear campaigns by competitors because behind it lies genuine anxiety about job replacement.

Nowadays, driving for ride-hailing services and delivering food have become fallback employment options for "laid-off" workers. As a significant reservoir of employment opportunities, the ride-hailing industry has absorbed a large number of workers.
Data from the Ministry of Transport shows that as of the end of December 2023, the number of registered ride-hailing drivers reached 6.572 million, surpassing 6.5 million for the first time. In 2023 alone, the number of new ride-hailing drivers was as high as 1.482 million.
Since the beginning of this year, many places across the country have issued risk warnings for the ride-hailing industry, publicly warning that ride-hailing capacities in many areas have become saturated.
According to the latest statistics from the Guangzhou transportation department, daily average earnings for ride-hailing drivers in Guangzhou have dropped from 343.34 yuan to 311.63 yuan. This means that even if ride-hailing drivers work "non-stop for the whole month," their monthly income does not exceed 10,000 yuan (Baiguan: ~1,375 USD) without deducting vehicle costs.
With the ride-hailing industry already so saturated, allowing Apollo Go Robotaxi to enter the market will undoubtedly anger many ride-hailing drivers.
For a long time, many people have opposed advancing AI primarily because "AI will replace humans." The scenario described in "Homo Deus" — "in the future, the vast majority will become useless people replaceable by AI, while a tiny minority will become gods who control AI" — contradicts the human-centered principle and has become the focal point of AI value alignment.
Many believers in AI say, "AI will not replace humans but will become a tool and assistant to humans; it will not replace people but will replace those who do not use AI."
However, with autonomous taxis hitting the road, job replacement and personnel displacement are almost inevitable.
Apollo Go Robotaxi is expected to deploy a thousand units of the sixth-generation autonomous vehicles in Wuhan this year. The impact of over a thousand Apollo Go Robotaxi vehicles on the ride-hailing market may be limited, but many people fear not the present but the future.
[ Baiguan: Screenshot of netizens discussing how low-cost robotaxi will impact human's livelihoods: In just one month, 1,000 cars completed 3 million orders, with an average of 100 orders per car per day. Apart from the time needed for charging, the autonomous taxis can operate 24 hours a day without needing breaks, rest, or drivers who might get tired. Additionally, there are no labor costs. The greatest advantage of autonomous taxis lies in their low cost and affordable prices. For example, for the same distance, a regular taxi might charge 20 yuan, whereas the autonomous taxi charges only 7 yuan. What would you choose? If autonomous taxis become profitable, the next step might see the emergence of autonomous delivery personnel and autonomous drivers for food delivery. Those currently working as ride-hailing drivers, taxi drivers, or food delivery personnel should consider transitioning early to adapt to the long-term changes. ]
According to my research, in Wuhan Economic Development Zone, Apollo Go Robotaxi's starting fee is 15-16 yuan, with additional charges for distance and duration.
However, some people online say that using Apollo Go Robotaxi is much cheaper than ride-hailing. Some say they only paid 4 yuan for a 6-kilometer ride, while others say they paid just over 1 yuan for 5 kilometers.
Reports say that in Wuhan, the fare for a 10-kilometer journey in an Apollo Go Robotaxi vehicle is 4-16 yuan, while in regular ride-hailing it is 18-30 yuan. "Autonomous taxis can save users nearly half the money."
Considering that autonomous taxis can accept orders 24 hours a day and operate around the clock, how many ride-hailing and taxi drivers can compete with them? Even if they were as tireless as oxen or horses, they wouldn't be able to keep up.
It's no wonder that some people are criticizing Apollo Go robotaxi with the argument that "the direction of high-tech development should be the stars and the sea, not competing with ordinary people for their livelihoods. [Baiguan: In the Chinese context, the "stars and the sea 星辰大海" is often used metaphorically to refer to grand and ambitious goals, such as exploring space.]
04
Some people hope Apollo Go Robotaxi can run faster, surpass Tesla, and crush Waymo; others hope it can run slower to wait for those at the grassroots who still rely on ride-hailing for their livelihoods.
So, how should we view the large-scale deployment of Apollo Go Robotaxi?
In my opinion, we shouldn't always try to slow down the pace of autonomous driving technology.
Kevin Kelly said: For problems caused by technology, we should never rely on reducing technology to solve them, but should invent better technology. Zhou Hongyi said: AI is still immature now, with inherent flaws and the possibility of being abused, it needs to be governed, but this cannot be a reason to hinder its development — for AI, not developing is the biggest insecurity.
The development of new technologies like autonomous driving will inevitably bring certain impacts, but we should not only see the negative externalities but also the positive ones.
Across the ocean, the United States is increasingly giving the green light to Tesla and Waymo's autonomous taxis, and China has no reason to press the pause button for autonomous taxis.
Online, some netizens find Apollo Go robotaxi very convenient: "No more awkward small talk, a blessing for introverts," "No more driver-passenger conflicts," and "No more worrying about not getting a ride on rainy days."
As long as Apollo Go Robotaxi can create value for users (such as increasing travel options, reducing travel costs, and eliminating travel obstacles), there is a reason for Apollo Go Robotaxi to exist. Because the market needs it.
05
[Baiguan: a screenshot of a netizen discussing the benefits of robotaxi: "Tried riding an autonomous taxi, 5 kilometers for only 1.15 yuan. Here are my honest feelings: 1. There is no driver in the car, and while I didn’t worry about safety, it felt strange, probably because it was my first time experiencing an autonomous vehicle. 2. Without a driver, I avoided awkward small talk, which is great for someone with social anxiety like me. 3. The entire journey was quite smooth. Autonomous driving technology is very mature, and to be honest, the passenger experience is not much different from having a human driver. 4. The smart screen in the back seat can play music and videos, and show road conditions, and the overall functions are simple to operate. 5. There's no need to worry about the driver canceling the order. Sometimes, regular drivers cancel due to distance issues, but with the autonomous taxi, the order is not canceled. 6. The price is really affordable. Although this puts pressure on ride-hailing and taxi drivers, from a consumer's perspective, as long as safety is ensured, why shouldn’t we opt for a more convenient and cheaper mode of transportation? ]
Automatic stopping on the roadside during heavy rain for an hour, slow speed... Currently, Apollo Go Robotaxi indeed faces various problems, but these problems can be solved as it develops.
However, in the current environment, Apollo Go robotaxi may find it difficult to use the "car vs. carriage" story to justify the social impact it has caused.
To some extent, it is Apollo Go Robotaxi's fate to be in a dilemma — What it originally had to endure was merely the game between e/acc (effective accelerationism) and EA (effective altruism), but the complex social structure has added variables to this game.
In the end, Apollo Go Robotaxi's idealized "AI vision" inevitably has to be embedded in the chain of "economic growth pressure - severe employment situation - tight employment reservoir - heightened anxiety about AI replacement," ultimately colliding with the wall of public resentment under the pull of harsh reality.
This is the "unjust accusations and resentment" that Apollo Go robotaxi is destined to endure: the deployment of autonomous taxis will inevitably stir up a lot of resentment, but Apollo Go robotaxi might also feel wronged—it has unintentionally become an outlet for social emotions concentrated on employment issues.
05
Apollo Go Robotaxi may run faster, but it cannot escape the "gravitational force" of reality.
Reality forces Apollo Go Robotaxi to "want both" — to bear the hope of China's leap forward in autonomous driving and to avoid attracting the anger of the ride-hailing driver group.
If conditions allowed, Apollo Go robotaxi would probably wish to push the large-scale deployment back to 2015—at that time, new business models could occupy the high ground with the popular narrative of "car vs. carriage," gain public support through the "new economy filter" of the masses, and benefit from the many entrepreneurial and employment opportunities available then, so people wouldn't flock to the ride-hailing industry.
But, this can only be "if."