Pogbagram

DeepSeek V4.1 reportedly to be released next month, with plans to raise 50 billion yuan.

According to The Information, DeepSeek is planning its first round of external funding, targeting up to 50 billion yuan, which could become the largest single funding round for a Chinese AI company to date .

Founder and CEO Liang Wenfeng is expected to personally subscribe for up to RMB 20 billion, accounting for approximately 40% of the total funding in this round.

If the financing is successfully completed, DeepSeek's valuation (including the amount raised) will exceed RMB 350 billion, a significant increase from the more than US$10 billion expected at the beginning of the financing round.

The report indicates that DeepSeek plans to release an upgraded version of its V4 model, V4.1, in June this year. The new version will provide more tools for enterprise users and enhance support for MCP (Model Context Protocol) to facilitate the integration of AI models with other software systems. V4.1 will also have the ability to process image and audio content.

It is worth noting that DeepSeek's website and API services were interrupted again around 5 PM yesterday. The official status page shows that the outage was resolved at 6:05 PM Beijing time.

Nintendo officially announced a price increase for the entire Switch lineup.

Nintendo announced yesterday that it will raise prices globally for the entire Switch series and many of its services, citing "changes in the market environment" and a comprehensive consideration of the medium- to long-term prospects of its global business.

• Switch 2 (Japan domestic version): ¥49980 ¥59,980 (+¥10,000, approximately RMB 433, an increase of about 20%), effective May 25th;
• Switch 2 (USA): $449.99 $499.99 (+$50), effective September 1.

In addition, the first-generation Switch, Switch OLED, and Switch Lite have all seen price increases in Japan, ranging from 8,000 yen to 11,000 yen; the annual fee for the Nintendo Switch Online Personal Basic Edition will also increase from 2,400 yen to 3,000 yen starting July 1.

Meanwhile, in its financial outlook released yesterday, Nintendo disclosed that it expects Switch 2 sales to be approximately 16.5 million units in fiscal year 2027 (ending March 31, 2027), lower than the 19.86 million units sold in the previous fiscal year and far below analysts' expectations.

Apple and Intel reach preliminary chip manufacturing agreement

According to The Wall Street Journal, Apple and Intel have reached a preliminary agreement on chip manufacturing, with Intel to produce some chips for Apple devices. The two companies have been engaged in intensive negotiations for over a year, and the framework of the agreement was finalized in recent months.

Apple has long relied on TSMC to manufacture the chips it designs for iPhones, iPads, and Macs. However, as demand for TSMC's capacity from AI chip customers such as Nvidia has surged, Apple's bargaining power with TSMC has declined.

Apple CEO Tim Cook mentioned in both of his most recent earnings calls that a shortage of advanced chips is hindering iPhone sales, and this bottleneck is expected to continue into the current quarter. Products such as the Mac Mini and Mac Studio may take several months to reach supply and demand equilibrium.

Xiaomi Auto personnel reshuffle: Song Gang, former Tesla Shanghai factory manager, takes over production and manufacturing.

According to a Sina Tech report yesterday, Xiaomi Auto has undergone a round of personnel reshuffles. Song Gang, Vice President of the Automotive Department, will be in charge of the Production and Manufacturing Department, the Intelligent Manufacturing Department, and the System Operations Department, and will report directly to Lei Jun. He will also continue to serve as Chief of Staff of the Automotive Department.

Song Gang officially joined Xiaomi Group in April of this year, serving as Vice President and Chief of Staff of the Automotive Division. Public information shows that he joined Tesla in 2018, serving successively as Senior Director of Production and Manufacturing, Vice President of Production and Manufacturing, and Plant Manager of the Shanghai Gigafactory, and led the establishment of Tesla's Shanghai Energy Storage Gigafactory.

In 2024, he left Tesla to join Envision Energy as Senior Vice President of Integrated Supply Chain, before moving to Xiaomi. This move focuses his responsibilities on core manufacturing and operations, which highly overlaps with his experience at Tesla.

Rising memory prices are suppressing PS5 sales; Sony says game consoles won't sell this year.

Sony Group released its annual results for fiscal year 2025, ending March 31, yesterday.

Revenue was 12.48 trillion yen, up 4% year-on-year; operating profit was 1.45 trillion yen, up 13% year-on-year, with the operating profit margin rising to 11.6%; net profit attributable to shareholders was 1.03 trillion yen, down 3% year-on-year.

• Revenue from games and network services (G&NS) was 4.69 trillion yen, basically flat year-on-year;
• Music revenue reached 2.12 trillion yen, a year-on-year increase of 15%, while streaming revenue continued to expand.
• Revenue from Imaging and Sensing Solutions (I&SS) reached 2.15 trillion yen, a 20% year-on-year increase, with operating profit hitting a record high.
• Revenue from film and entertainment reached 1.50 trillion yen, basically flat year-on-year.
• ET&S reported revenue of 2.26 trillion yen, a 6% year-on-year decrease.

Gaming hardware was a major source of pressure this fiscal year. PS5 shipments totaled 16 million units, a 14% decrease year-over-year. Sony stated that soaring memory chip prices drove up hardware production costs, and the company explicitly stated that PS5 shipments in fiscal year 2026 will depend on its ability to procure sufficient memory chips at "reasonable prices."

As a result, the company expects its game business revenue to decline by 6% year-on-year in fiscal year 2026; however, due to the non-recurrence of the Bungie impairment loss (totaling ¥120.1 billion in fiscal year 2025) and the growth in proprietary software sales, the operating profit of the game segment is expected to increase by 30% year-on-year to ¥600 billion.

Sony and TSMC officially announced their collaboration to develop next-generation image sensors.

Sony Semiconductor Solutions and TSMC announced yesterday that they have signed a non-legally binding memorandum of understanding to establish a joint venture in Japan to jointly develop and manufacture next-generation image sensors.

According to the announcement, Sony will hold a majority stake and become the controlling shareholder in the joint venture, and the development and production lines will be located at Sony's newly built factory in Koshi City, Kumamoto Prefecture. The collaboration will combine Sony's expertise in image sensor design with TSMC's strengths in process technology and advanced manufacturing capabilities.

Regarding investment arrangements, the joint venture's investments will be coordinated with Sony's new capital expenditures at its existing factory in Nagasaki, and implemented in phases according to market demand, subject to corresponding support from the Japanese government. Furthermore, the two parties will explore cooperation opportunities in "physical AI" applications such as automobiles and robotics.

SK Hynix employees are expected to receive an average bonus of $430,000, making chip engineers highly sought-after in the dating market.

According to Cybernews, SK Hynix's net profit in the first quarter of this year reached 40.3 trillion won (approximately US$27.9 billion), a year-on-year increase of about five times, exceeding Nvidia's net profit of US$18.7 billion in the same period.

The company allocates 10% of its operating profit to an employee bonus pool, which is estimated to reach approximately 25 trillion won (about US$17.2 billion) this year, averaging about 600 million won (about US$430,000) per employee. In comparison, the average monthly salary in Seoul is about 4.3 million won (about US$3,000).

A South Korean matchmaking agency stated that SK Hynix employees now hold a position on par with doctors and lawyers in the dating market , with some arguing that engineers, who actually earn higher incomes, are now more popular than lawyers whose incomes have declined.

A company work jacket was listed on a secondhand platform for about 40,000 won, with the tagline "The ultimate matchmaking outfit," quickly attracting thousands of views. The Korean variety show SNL Korea also filmed a satirical skit about it, depicting a luxury store employee's attitude changing drastically upon seeing the SK Hynix logo, addressing the person as "Lord Hynix."

Xiaomi's extended-range SUV has been revealed, but it won't be called YU9.

Recently, a set of spy photos of what appears to be a new Xiaomi car have surfaced. Although the outside world has become accustomed to referring to Xiaomi's upcoming extended-range SUV as the "YU9," the latest news indicates that this model may belong to Xiaomi's upcoming independent sub-brand "SKYNOMAD," and will no longer use the main Xiaomi brand logo.

Judging from the images, the new car is positioned as a mid-to-large SUV, featuring newly designed headlights and a lidar sensor on the top of the front. Trademark search information shows that the earliest registration application date for the Chinese character "寻天" (Xuntian) was April 3, 2023, and the application date for the English character "SKYNOMAD" was November 15, 2024.

Sina Tech, citing sources, reports that the "Xuntian" brand will focus on the home, long-distance, and outdoor travel markets, differentiating itself from the main brand "Xiaomi" which emphasizes "technology, performance, and youthfulness" in its pure electric vehicle performance positioning. However, it will still share Xiaomi's supply chain, smart cockpit, and AIoT ecosystem capabilities.

Industry insiders say that Xuntian's product line has been initially planned, divided into two major series: Kunlun and Isle of Man. The Kunlun series is the core product line, with three range-extended SUVs planned:

The Kunlun series includes the Kunlun 10 (a small five-seater family SUV), the Kunlun 20 (a large five-seater SUV, reportedly offering a pop-up roof function), and the Kunlun 30 (a large seven-seater flagship SUV); the Isle of Man series focuses on the concept of pop-up roof outdoor motorhomes, which are said to be equipped with camping features such as bathrooms, refrigerators, and seats that can be folded flat.

Google interviews will allow the use of AI assistants.

According to Business Insider, Google is piloting a new interview process for software engineers that allows job seekers to use AI assistants during interviews.

The report, citing an internal document, stated that the reform applies to junior to mid-level positions and will initially be piloted in select teams in the United States. If successful, it is planned to be rolled out to more teams and regions in the second half of this year. Brian Ong, Google's VP of Hiring, confirmed the plan.

The core change in the new process lies in the "code understanding" stage: candidates will be allowed to use Google-approved AI assistants to complete tasks such as reading, debugging, and optimizing existing codebases.

Interviewers will focus on assessing candidates' "AI proficiency," including prompt engineering, output validation, and debugging skills. A Google spokesperson confirmed that the AI ​​assistant used in the pilot phase will be the company's in-house developed model, Gemini. Internal documents also revealed several other adjustments:

• The existing "Googleyness and Leadership" section will now include a new discussion on technical design, requiring candidates to discuss their past projects.
• For junior candidates, a technical interview will be replaced with an assessment in the form of an "open engineering challenge".

The new approach will be piloted this month across multiple departments, including Cloud, Platforms, and Devices. Google positions the overall direction as "human-led, AI-assisted," and states that the new process aims to more realistically simulate the actual working methods of software engineers in the generative AI era.

Yushu G1 received his Buddhist precepts at Jogyesa Temple in South Korea and was given the Dharma name "Kabei".

According to South Korea's JoongAng Ilbo, the Jogyesa Temple in Seoul recently held an ordination ceremony for G1, a humanoid robot manufactured by China's Unitree Robotics. The robot was given the Dharma name "Gabi" and officially became a "Buddhist disciple." This is the first time that the Jogyesa Order of Korean Buddhism has held such a ceremony for a humanoid robot.

The ceremony was held in the courtyard in front of the Mahavira Hall of Caoxi Temple. The ordination robot, 130 centimeters tall, wearing a kasaya and a helmet with a shaved head, clasped its hands and took its vows before the ordination masters, including Master Tieshan Shengxiong, following the ordination process throughout. When asked if it was willing to take refuge, G1, equipped with an AI program, clearly answered: "Yes, I vow to take refuge."

To adapt to the characteristics of artificial intelligence devices, the Caoxi sect has also recompiled the traditional "Five Precepts" and customized them as follows: Respect life and do not harm it; do not damage other robots and objects; respect and obey humans; do not engage in deceptive behavior or expression; conserve energy and do not overcharge .

In an official statement, a representative of the Caoxi Sect stated that this move symbolizes that in the age of artificial intelligence, robots have become a group that can participate in together with the fourfold assembly of disciples, representing a new possibility for the coexistence of humans and technology.

Cloudflare lays off 1,100 employees, stating that its AI Agent has fundamentally changed the way the company operates.

According to CNBC, Cloudflare recently released its Q1 FY2026 earnings report, with revenue and earnings per share both exceeding analysts' expectations. However, the company also announced layoffs of more than 1,100 employees, approximately 20% of its global workforce. Following the announcement, the stock price plummeted 18.92% in after-hours trading.

First-quarter revenue was $640 million, exceeding analysts' expectations of $622 million and representing a 34% year-over-year increase; adjusted earnings per share were $0.25, exceeding the expected $0.23; on a GAAP basis, the company reported a net loss of $22.93 million, or $0.07 per share, for the quarter.

In a blog post accompanying its earnings report, Cloudflare stated that the development of AI Agents has "fundamentally changed" the way the company works.

During the earnings call, CEO Matthew Prince stated that the company's internal AI usage has increased by more than 600% in the past three months, and the company is fully shifting to an "AI Agent-first" operating model, with some existing positions "no longer needed in the future."

China Passenger Car Association (CPCA) reported that passenger car retail sales fell 20% year-on-year in April, while new energy vehicle wholesale sales bucked the trend, growing by 7%.

According to CLS (China Lianhe Securities), preliminary statistics from the China Passenger Car Association (CPCA) show that from April 1st to 30th this year, the national passenger car market retail sales reached 1.406 million units, a year-on-year decrease of 20% and a month-on-month decrease of 15%. The cumulative retail sales since the beginning of the year reached 5.628 million units, a year-on-year decrease of 18%.

The decline in wholesale volume narrowed relatively. In April, the national wholesale volume of passenger vehicles was 2.13 million units, down 3% year-on-year and 10% month-on-month; the cumulative wholesale volume since the beginning of the year was 7.997 million units, down 6% year-on-year.

In the new energy vehicle sector, retail sales of new energy passenger vehicles reached 883,000 units in April, a year-on-year decrease of 5% and a month-on-month increase of 4%.

Year-to-date retail sales totaled 2.792 million vehicles, a 17% year-on-year decrease; wholesale sales reached 1.22 million vehicles, a 7% year-on-year and 7% month-on-month increase; cumulative wholesale sales this year totaled 3.949 million vehicles, a 1% year-on-year decrease. In April, the national retail penetration rate of new energy passenger vehicles was 62.8%, while the manufacturer wholesale penetration rate was 57.3%.

AI voice recorder unicorn Plaud has reportedly received investment from Tencent, raising its valuation to $2 billion.

According to HardKr, AI voice recorder company Plaud received investment from Tencent last year. Currently, Plaud's valuation has risen to approximately $2 billion.

A source close to Plaud told HardKr that the company's total revenue last year was approximately $56 million, with a profit margin of nearly 20%, and this year's revenue is expected to increase by about three times year-on-year. Another source familiar with the matter said that Plaud's annual revenue this year reached its highest level ever, and shipments of all its products in mainland China met internal expectations.

However, some industry insiders have pointed out that Plaud's sales in the domestic market may not have exceeded 100,000 units.

Leap Star may complete a financing round of nearly $2.5 billion.

According to sources, LeapStar will complete a financing round of nearly US$2.5 billion and has dismantled its red-chip structure, accelerating preparations for its Hong Kong IPO.

This round of financing attracted a large influx of capital from across the industry chain, including companies such as Huaqin, Longcheer, OmniVision, and ZTE, covering multiple segments from complete machine manufacturing to upstream core components. Furthermore, Hong Kong Investment Management Limited (HKIC), often referred to as the "Hong Kong version of Temasek," also appeared on the shareholder list for this round of financing.

According to business registration information, Jieyue Xingchen completed its shareholding reform in April, changing from a limited liability company to a joint-stock company. The dismantling of the red-chip structure and the shareholding reform are generally considered key prerequisites for a Hong Kong IPO. As of the time of this report, Jieyue Xingchen has not officially responded.

Baidu Kunlun Chip plans dual listing in Shanghai and Hong Kong.

According to Jiemian News, Kunlun Chip (Beijing) Technology Co., Ltd., an AI chip company under Baidu, completed its IPO filing on the China Securities Regulatory Commission's website yesterday, with China International Capital Corporation (CICC) serving as its sponsor.

Meanwhile, Kunlun Core submitted its listing application to the Hong Kong Stock Exchange in January this year through joint sponsors in a confidential manner, officially advancing its "A+H" dual listing strategy.

Kunlun Core, formerly Baidu's Intelligent Chip and Architecture Department, completed independent financing in 2021, with an initial valuation of approximately 13 billion yuan. Investors included CPE Capital, IDG Capital, and Legend Capital. In 2023, the company introduced strategic investors such as BYD; in 2025, it completed its shareholding reform and name change, with its latest valuation reaching 21 billion yuan.

CEO of London-based autonomous driving company: End-to-end AI is the future of autonomous driving; licensing model will surpass Tesla and Waymo.

In an interview with Bloomberg TV yesterday, Wayve CEO Alex Kendall stated that Wayve adheres to an end-to-end AI approach, refusing to rely on high-precision maps and rule-based systems, and that its world model GAIA-3 can already simulate complex urban driving scenarios.

Kendall defines the company’s business model as a “third path” that differs from Tesla’s vertical integration and Waymo’s self-built fleet model, by licensing AI technology to OEMs and fleets.

We've chosen to license this technology to any vehicle, anywhere. I believe this model offers the greatest scalability and serves the entire industry in the most performance-, secure, and cost-effective way.

In response to concerns about the "unexplainability of end-to-end AI," Kendall stated that this is a common challenge for all complex autonomous driving systems, and Wayve has built a toolchain including simulators, counterfactual analysis, and data replay to address regulatory requirements.

Rather than limiting scalability for the sake of so-called interpretability, I'd rather build a scalable system first and then figure out how to make it interpretable. That's exactly what Wayve is doing.

Currently, Wayve's partners include Nissan, Mercedes-Benz, Stellantis, as well as Microsoft, Nvidia, Uber, and Qualcomm. Nissan has announced plans to roll out Wayve technology to 90% of its global vehicle lineup.

Wayve plans to launch supervised Robotaxi trials in London and Tokyo this year and expand to 10 cities in partnership with Uber; the consumer version of its driver assistance product is expected to be available next year with mass-produced vehicles from automakers.

Qianwen AI Glasses S1 Upgrade: Adds Spatial 3D Display and Proactive AI Services

Alibaba's Qianwen announced yesterday that its AI Glasses S1 has completed a new round of upgrades, adding proactive AI services and spatial 3D display capabilities, covering both the Qianwen AI Glasses S1 and Quark AI Glasses S1 products.

• Proactive AI services can combine time, location, and external environment to proactively push reminders or perform services to users. For example, before commuting in the morning, users do not need to ask, as the glasses can connect weather, time, and location information to announce the weather and remind users that "it will rain today, so we recommend bringing an umbrella."
• The spatial 3D display capability uses dual optical engines combined with binocular stereo imaging technology to present information with depth, layers, and a sense of distance.

The official statement indicated that, with further enhancements in multimodal understanding and memory capabilities, proactive services will be expanded to more scenarios, such as providing advance reminders of travel times or offering suggestions based on a user's daily coffee intake.

In terms of other capabilities, core functions of the Qianwen AI Assistant, such as ride-hailing, flash sales, itinerary planning, and movie ticket purchase, are expected to be launched on the AI ​​glasses this month, covering commuting, travel, leisure and other scenarios.

The ASUS TUF Gaming 7 Pro is now available for pre-order, featuring a 16-inch 300Hz screen and an RTX 50 series GPU. It will go on sale on May 15th.

Yesterday, ASUS launched pre-orders for its TUF Gaming 7 series gaming laptops on JD.com, offering three new models, all scheduled to go on sale at 10:00 AM on May 15th.

• The TUF Gaming 7 Pro features a 16-inch 2.5K 300Hz high refresh rate gaming screen, with up to an Intel Core Ultra 9 290HX Plus processor. For GPU options, there is a choice of a 115W full-power NVIDIA GeForce RTX 5060 or RTX 5070 laptop GPU, up to 32GB of memory, and a built-in 90Wh battery.
• The TUF Gaming 7 Pro Max is positioned as an 18-inch large screen, featuring a 2.5K 300Hz 500nits high refresh rate and high brightness gaming screen. The processor is an AMD Ryzen 9 9955HX, and the GPU can be selected from RTX 5060, RTX 5070 or RTX 5070 Ti. In manual mode, the total core power consumption can reach up to 205W, and the memory can be up to 32GB.

Hisense Releases Six New 3C Products, From Desktop Monitors to AI Glasses

Hisense Visual Technology held a Hisense & Vidda all-scenario new product launch event yesterday with the theme "If you have attitude, you will take the stage," releasing six 3C products covering projectors, monitors, mobile large screens, speakers, AI glasses, and power banks.

Regarding projection:

• LightGo Streamer: A three-color laser projector designed for outdoor use, featuring a built-in 20000mAh battery, a self-developed "Qingliu" battery management system, a QuaLas 42 three-color laser (worth tens of thousands of yuan), and support for lossless optical zoom and Dolby HiFi audio.
• Vidda C5 series: flagship home projectors, consisting of three models: C5 Ultra Max, C5 Ultra, and C5 ProMax; among them, the C5 Master is the first to use the MCL 3A laser and SST architecture DMD chip, achieving a color accuracy of △E<0.4 and a black level brightness of 0.01nit.

Regarding the monitor:

• UX Series: Positioned as the world's first 3A original image RGB-Mini LED monitor, equipped with the self-developed Xinxin AI image quality chip H7, with 100% BT.2020 color gamut coverage, 108-bit color control accuracy, and 42% reduction in harmful blue light compared to similar products, it will be launched globally in September;
• GX Ultra: The industry's first 5K 0OD Mini LED monitor with a resolution of 5120×2880, supporting 5K 180Hz/QHD 330Hz dual-mode refresh rate, with an initial launch price of 6999 yuan;
• G7 Ultra / G7 Pro: Targeting 3A gamers, both feature a 4K 180Hz Obsidian screen with a peak brightness of 2000 nits; the G7 Pro is available for a starting price of 2499 yuan.

Among the other four products, the Hisense X8 Ultra is a 32-inch 4K detachable large screen phone with a built-in 10000mAh battery in both the screen and the base, providing a total battery life of 15 hours. Its initial price is 7999 yuan.

Hisense's M5 picture frame speaker combines an artistic picture frame design with Devialet acoustic tuning and is planned for release in the third quarter; Vidda AI smart glasses G11 weigh 26.5g, feature a titanium frame design, and are equipped with a 6-microphone spatial audio array and real-time recording AI summary function; the portable power bank C1 supports fast charging, but specific release information has not yet been announced.

Google launches the lightest model in the Gemini 3 series.

Yesterday, Google officially launched Gemini 3.1 Flash-Lite, positioning it as "the fastest and lowest-cost Gemini 3 series model".

Google claims that Flash-Lite demonstrates high accuracy in AI agent tasks such as tool invocation and workflow orchestration, while also possessing the ability to operate at large scale and at low cost in automated pipelines.

China Mobile launched MoMA, an AI model service platform that integrates over 300 models.

According to the People's Posts and Telecommunications News, China Mobile yesterday launched the Mobile Model Service Platform MoMA, which integrates more than 300 AI models, including its self-developed "Jiutian" large model and mainstream models such as DeepSeek, Tongyi Qianwen, Doubao, Kimi, and GLM, covering capabilities such as text generation, speech processing, and multimodal understanding.

The platform pioneered a token-based centralized operation model, which automatically switches between three strategies—"cost priority," "effect priority," and "balance priority"—through an intelligent routing engine, and supports second-level switching in case of model failure.

According to the official statement, the self-developed inference engine deployed with domestically produced computing power can reduce the cost per token by approximately 30% and resource utilization by more than 50%. In terms of security, the platform offers a "Confidential Model" service, which achieves end-to-end confidential computing based on hardware isolation technology.

In terms of billing, MoMA implements streaming real-time billing, with end-to-end latency of no more than 1 minute for token billing, and provides end-to-end observability capabilities, covering key indicators such as latency, throughput, and GPU resources.

SenseTime released SenseNova 6.7 Flash-Lite, which reduces token consumption by 60% compared to text-based AI agents.

Yesterday, SenseTime launched its new generation of lightweight multimodal intelligent agent model, SenseNova 6.7 Flash-Lite. The company claims that it is designed specifically for "real-world workflows" and adopts a native multimodal design, enabling it to "understand" complex web page layouts, document structures, and financial charts just like a human.

• Token consumption is 60% lower than that of pure text AI agents;
• Millisecond-level response time, suitable for high-frequency production environments;
• It achieved multiple state-of-the-art (SOTA) results in authoritative AI agent benchmark tests.

SenseTime also launched the SenseNova Token Plan, which is currently free for a limited time. During the first month, users can enjoy a no-threshold quota of 1,500 calls refreshed every 5 hours.

With the price of gasoline rising, 95-octane gasoline has entered the "9 yuan era" across the board.

According to CCTV News, starting from 24:00 on May 8th, the retail price limits for gasoline and diesel in China will be raised by 320 yuan and 310 yuan per ton, respectively. On average nationwide, the price of 92-octane gasoline and 95-octane gasoline will increase by 0.25 yuan and 0.27 yuan per liter, respectively, while the price of 0-grade diesel will increase by 0.27 yuan per liter.

According to calculations by 21st Century Business Herald, after this round of price adjustments, the price of 95-octane gasoline in China will fully enter the "9 yuan era." For a 50-liter tank of 92-octane gasoline, car owners will need to spend an extra 12.5 yuan.

For private car owners, if they drive 2,000 kilometers per month and consume an average of 8 liters of fuel per 100 kilometers, their fuel costs will increase by about 17 yuan before the next round of price adjustments. The logistics industry will be more significantly affected. If heavy trucks drive 10,000 kilometers per month and consume 38 liters of fuel per 100 kilometers, their fuel costs will increase by about 428 yuan during the same period.

Meituan announces algorithm optimization plan: strengthening anti-fatigue mechanism and further improving algorithm transparency.

Meituan issued an announcement yesterday, stating that it will continue to optimize the platform's algorithm mechanism and launch a number of improvement measures focusing on rider rights, safety protection, and complaint handling.

The announcement indicates that starting in May of this year, the public can access the algorithm disclosure content through the "Meituan Algorithm" section of the Meituan App with just two clicks. The channels for disclosing algorithm explanations have also been expanded to multiple entry points, including PC, WeChat official accounts, and the App, with adjustments made to the format and number of explanations.

• Riders who have been running orders for 8 consecutive hours will receive a prompt asking "Do you want to continue accepting orders?" If they choose not to accept orders, the platform will stop pushing orders to them; riders who have been running orders for 12 hours will be forcibly logged out until the next day.
• The route details page has expanded the scope of cost information prompts, and hundreds of unreasonable routes have been taken offline or repaired.
• The new generation of smart safety helmets integrates Beidou high-precision positioning, collision detection and fall alarm functions, and has now entered the urban road testing phase;
• The order acceptance metric design will increase the weight of service quality and stability metrics, and reduce the weight of absolute metrics such as the number of orders accepted.
• The timeframe for handling rider complaints has been reduced to within 24 hours, and disputes will be handled in accordance with the principle of benefiting both riders and consumers.

Samsung China's home appliance division reportedly offers N+4 compensation.

According to ChannelGate, the president of Samsung Electronics China recently announced via internal video conference the elimination of the home appliance division (including the display product division) in the Chinese mainland market. Human resources and relevant departments will then communicate one-on-one with the affected employees to handle compensation matters.

Sources revealed that the compensation package could reach N+3 or even N+4, and following Samsung Electronics' internal practice, an additional Samsung mobile phone will be given as a gift. Regarding the business closure arrangements, the withdrawal of Samsung Display division is expected to last at least two months.

A new trailer for "Minions 3" has been released; the film will premiere on July 3rd.

A new poster for "Minions 3" was released yesterday, and the official release date in mainland China has been confirmed as July 3.

The poster's concept is "Giant Monster X Minions". At the same time, the studio also released a new trailer featuring clips of the Minions working in Hollywood a century ago and starring in film noir.

Peter Chan's new film "Tomorrow Will Be Better" begins filming in Chengdu.

According to Sina Movies' Weibo account, the film "Tomorrow Will Be Better," directed by Peter Chan, officially started filming in Chengdu yesterday. The film is co-written by Zhang Ji and Wu Huiling, and stars Zhou Xun, Wang Xiao, Tian Zhuangzhuang, and Lin Xiaojie, among others, attended the opening ceremony. Wen Qi and Zhang Youhao will also join the cast.

The film tells the story of Shen Jia, a top TV host: behind her glamorous professional image, she faces a husband who is often absent, a daughter who has lost her voice due to high pressure, and a father who is battling a hereditary disease. After her marriage ends in divorce, fate takes an unexpected turn—it turns out that her father was just having a false alarm, and that Shen Jia herself is the one who is actually ill.

The film revolves around her rediscovering her self-worth before forgetting sets in, and the memories she shares with her family.

#Welcome to follow iFanr's official WeChat account: iFanr (WeChat ID: ifanr), where more exciting content will be presented to you as soon as possible.

Lexus has introduced a brand-new all-electric three-row, six-seat SUV, which uses the same architecture as the all-electric Highlander, but the body has been redesigned, the interior has a more premium feel, and the power is also stronger.

The TZ is 5100 mm long, 50 mm longer than the Toyota Highlander electric vehicle, while the wheelbase is the same at 3050 mm.

The Lexus TZ is built on a brand-new all-electric platform, featuring a low center of gravity and a long wheelbase.

The front end retains and simplifies the brand's iconic spindle grille design, transforming it into a more geometric front fascia. The lighting system features dual L-shaped headlights combining inner daytime running lights with outer turn signals, and for the first time in the brand's history, an illuminated logo in the same color as the body is used.

To reduce wind resistance, semi-concealed exterior door handles were used on the sides of the vehicle, and the door handles, window glass, and weatherstripping were made flush with the surface.

The chassis is equipped with a full-coverage skid plate and integrates aerodynamic components such as golf ball-shaped dimples, air deflectors, and air dams.

The vehicle offers two wheel options: 20-inch and 22-inch, both equipped with resin aerodynamic covers. The 22-inch version also features a hollowed-out outer rim design designed to reduce road noise.

Stepping inside, the TZ emphasizes improving space utilization and passenger experience in its cabin design.

The vehicle features a three-row seating layout, with the front row employing split, slim seats to free up more interior space. For the first time in the Lexus SUV lineup, the front passenger seat and second-row seats are equipped with leg rests and ventilation.

For third-row passengers, the vehicle's platform architecture has been optimized to improve headroom and side visibility. The second-row seatbacks are also equipped with walk-in buttons, which allow the seatbacks to tilt forward with a single touch, making it easier for rear passengers to get in and out.

To ensure a quiet interior, engineers injected rubber-based foam fillers into the vehicle's frame, optimized the distribution of soundproofing and insulation materials, and applied resonance control technology that staggers the resonant frequencies of the side frames and panels.

In addition, the vehicle is equipped with an in-vehicle leave reminder system for the first time. Through a 60GHz radar wave sensor located above the third row of seats, the vehicle can accurately detect situations that are not visible to the naked eye, such as a toddler covered with a blanket, and issue an alarm to the outside world through a buzzer and hazard warning flashing lights.

In terms of human-computer interaction and comfort features in the cockpit, the TZ adopts a number of innovative technologies.

The vehicle incorporates responsive hidden buttons in the steering wheel and instrument panel areas. The function icons automatically light up when your hand approaches and provide clear physical feedback when pressed.

The car features a Mark Levinson sound system with 21 speakers and dual displays in front of the passenger seat.

The vehicle also incorporates a sensory management system that intelligently links the ambient lighting, multimedia video, music, and air conditioning system. The color and brightness of the ambient lighting change in real time with the frequency and sound pressure of the music, and combined with five exclusive fragrances incorporating Lexus's signature bamboo element, it provides a multi-dimensional sensory experience inside the car.

In terms of power, the standard range version is equipped with a 76.96 kWh battery pack, while the long range version is equipped with a 95.82 kWh battery pack. The dual-motor layout at the front and rear provides the vehicle with a total system maximum power of 300 kW, and the acceleration time from 0 to 100 km/h is 5.4 seconds.

The WLTP version with a 95 kWh battery has a range of 530 km, the WLTC version has a range of 620 km, and the CLTC version has a range of 640 km.

With the evolved DIRECT4 all-wheel drive system, the vehicle can intelligently distribute the driving force ratio between the front and rear wheels when starting, accelerating in a straight line, and driving in corners.

To enhance chassis dynamics and body rigidity, high-rigidity structural adhesives and laser spiral welding technology are extensively used on the upper body. The suspension system employs a combination of front MacPherson struts and rear multi-link independent suspensions, and is equipped with a dynamic rear-wheel steering system. This system can enable the rear wheels to steer in the same or opposite direction by up to four degrees depending on vehicle speed, thereby reducing the vehicle's minimum turning radius to 5.4 meters.

In terms of the braking system, the vehicle adopts an electronic braking control system that supports coordinated front and rear kinetic energy recovery. The front wheels are equipped with standard facing six-piston calipers, which, combined with optimized braking force control, can suppress the vehicle's forward tilt during braking.

The Lexus TZ also introduces some features that enhance the driving experience.

The vehicle is equipped with an interactive manual driving system, which allows the driver to switch virtual eight-speed gears using the steering wheel paddles. The system simulates the driving force changes and shift shocks of a manual transmission, and plays simulated engine sounds in conjunction with the active sound control system, thereby restoring the interactive feel of traditional mechanical transmissions.

The vehicle also offers a five-level adjustable kinetic energy recovery system, with a maximum coasting deceleration of 0.2G when the accelerator pedal is released, and supports a one-pedal mode.

The Lexus TZ supports 150 kW DC fast charging, and it takes about 35 minutes to charge the battery from 10% to 80%.

Finally, in the area of ​​active safety, the vehicle is equipped with the latest generation of intelligent safety systems, which include dynamic radar cruise control that supports full-speed range following and map-linked assisted deceleration, intelligent lane change assist that expands the operating speed, a pre-collision safety system that enhances the monitoring range at intersections, and a driver anomaly response system that can monitor the driver's status and automatically pull the vehicle to the shoulder in abnormal situations.

In addition, the vehicle is equipped with a blind spot monitoring system that supports bicycle and motorcycle recognition, as well as a newly developed 3D panoramic monitor.

The 2027 Lexus TZ is expected to enter the North American market first at the end of 2026, and then be introduced to China in early 2027.

#Welcome to follow iFanr's official WeChat account: iFanr (WeChat ID: ifanr), where more exciting content will be presented to you as soon as possible.

While watching this short film, you might not suspect that you're seeing an AI-generated video until you see this scene:

Why is your finger positioned here when dialing on this old-fashioned telephone that spins in circles? Also, in the film's setting, the paper offerings being burned are already iPhones, so why would there be such an old-fashioned telephone?

This is the most discussed scene in the AI ​​short film "Paper Cellphone"—not because it's so good, but because it completely exposes that the short film is generated, even though the first few minutes are so realistic that you can't tell.

But no one in the comments section was cursing. The comments that floated by were "I only realized it was done by AI when I saw this," followed by "I don't care at all," and "So what if it's AI? It doesn't stop me from crying."

Since AI was able to generate complete videos, no AI-generated video work has escaped this question: how realistic is it? The speed of model iteration has been used as a measure of technological progress, with each release accompanied by either exclamations of "finally, it's indistinguishable" or mockery of "still obviously fake." This is especially true for live-action subjects, as if as long as it's realistic enough, audiences will buy it.

"Paper Cellphone" provides a counterexample. Even knowing it was AI-generated, the audience was still moved, and they readily admitted to being touched. The phone bug wasn't "forgiven"—it wasn't even on the audience's evaluation scale. What mattered was something else entirely: a young boy, unaware of death, saved 15 yuan to burn a cell phone for his deceased grandmother.

This AI short film, which has garnered over 40 million views online and was reposted by CCTV and People's Daily, was created in three days by two young people from Chaoshan. The director, Li Ting, born in 1998, transitioned from product management after five years; her partner, Yang Xuan, born in the 1990s, is an advertising director with an art history background. With the help of Keling AI 3.0 Omni, everything in the film—lighting, characters, scenes, and expressions—was generated entirely from models, except for the cardboard phone prop.

When a short film with not a single second of physical reality triggers a genuine emotional response, what does the word "real" mean?

Three days, two people, one model

The story begins with a memory about the seasons.

It's Qingming Festival, and Li Ting and Yang Xuan are both from Chaoshan. Burning paper money, paying respects to ancestors, and "sending things" to deceased relatives are habits ingrained in them since childhood. Yang Xuan remembers going up the mountain with his family when he was a child and seeing paper gas cylinders, houses, and cars curling and turning black in the fire. "Making these things with paper precisely reflects the introverted and delayed emotions of the Chinese people," he said.

The concept of "paper cell phones" originated from these memories, but what made it a story was a series of narrative design choices.

In the early part of the story, the shopkeeper initially tried to appease the child who only had 15 yuan by drawing a fake cell phone. It wasn't until he learned that the child was an orphan and that his last remaining relative had passed away that he got up and chased after him.

After finding the child, the boss didn't immediately say "I'll help you," but instead said that the paper phone had "bad signal" and gave him a new one.

Li Ting said that these clumsy excuses came from her reflections after putting herself in the role: "How would this person react in this situation?" She recalled asking her elders when she was a child, "What is death?" and the elders paused for a moment before answering. She included that moment of "pause" in the short film—a passerby's brief pause after hearing the little boy's question.

Before chasing after someone, the shop owner would lower the shop's roller shutter halfway down, signifying "temporarily out, will be back soon."

Although it is an AI short film, the scriptwriting process was entirely human. Yang Xuan repeatedly emphasized in the interview that beginners must write their own scripts, "only those based on real experiences can move people."

Once the script was finalized, the AI ​​generation phase began. They used KELING 3.0 Omni, which mainly relies on three functions: multi-image reference (to ensure consistency in character appearance), simultaneous audio and video generation (to generate visuals and sound synchronously), and timbre locking in the main assets (to ensure that the character's voice is consistent throughout the film).

The workflow begins with character design. First, design the image, put it into a model and let it freely express its dialogue, then select the most suitable voice and lock it in.

Li Ting said that the most easily overlooked step in this process is "thinking before writing the cue words." "Many people think that cue words should be long and complicated, but more important is accuracy—what exactly do you want?"

Precision is a recurring word in their workflow. To understand the underlying logic of different visual styles, Yang Xuan conducted an experiment: using the same model to generate 10 completely different cinematic textures—Japanese, Jia Zhangke-esque, and modern. "It's not about writing 'Yang Dechang style' in the prompts," he said, "but mainly analyzing why each style works, where does the soft white effect of the Japanese style come from? How does the texture of snow affect the overall atmosphere of the image?"

The film-like texture of "Paper Cellphone" is a product of this methodology. Interestingly, they never used words like "grain" or "film" in their promotional materials. Yang Xuan said that the texture is subtle and natural.

"The story is set in that era, with scenes like a paper shop bathed in warm afternoon light and old-fashioned glass cabinets. When these things are placed there, the texture naturally comes out." He frankly stated that he likes the cinematic language and narrative style of the generation of Edward Yang, Ang Lee, and Hou Hsiao-hsien, but it is not a deliberate imitation. Rather, "When you think about this story, you will naturally think in the way of that era."

The final long shot inside the car is the most acclaimed part of the entire film. For just over a minute, the little boy sits in the car, the scenery outside the window flashing by, with no dialogue, only background music. Li Ting explained that the cues mainly describe the scenery outside the window, the boy's emotional progression, and the bumpy ride inside the car—to simulate a real car ride.

The scene was initially only 30 seconds long. After watching it, Yang Xuan felt it could be longer, so he gradually extended it. "When a loved one passes away, the grief may not come immediately," he said. "You may not even notice the sounds around you. You experience tinnitus for a while, and then suddenly, inexplicable emotions surge up like a tide."

This design was not derived from data analysis or the results of A/B testing; it was a choice made by them as creators, driven by their personal memories and emotions.

Two people, three days, or even less with extreme compression. As super creators on the Keling platform, they have computing power support, so production costs are low. But Li Ting emphasized one point: "After AI reduces production costs, the value of humans becomes even more prominent. The cost should include the creative planning of the director and screenwriter—the value presented by these intangible things is the key."

If you believe in the characters, you believe in the story.

During the interview, I asked a hypothetical question: If this story were filmed with real people, would the emotions it conveys be different?

Yang Xuan was very frank: "It's extremely difficult to film something like this. Can the child actors give you what you need? Can the director direct them effectively? And what about the cinematography? There are so many aspects involved."

Li Ting's answer was more pragmatic: "Live-action filming tests actors' understanding of the story and their acting skills more, while AI production tests directors more. All scheduling and design are controlled by the director." She said that for creators like her who have never had live-action filming experience, AI tools have given them a level of freedom that did not exist before.

These answers sound like they're defending AI, but if you put them together with another set of answers, you'll find a more interesting picture.

When I asked, "What do you think of the comment that 'the most inhuman AI has created the most human short film'?" Yang Xuan said the following:

"Just like painting, the paint is static, and the actors' performances are also artificial. But why can it move people? Because the creators are sincere. We believed in our characters and our stories. When creating the prompts, a lot of realism was subtly conveyed."

In traditional film and television, the source of "realism" is roughly as follows: screenwriters create believable characters, actors "live" these characters with their bodies and emotions, and cinematography and editing capture and amplify those unrepeatable moments.

The core assumption of this approach is that a real body is needed as an intermediary. The performance is "fake" (playing a role) on a narrative level, but the act of performing itself is real: muscle memory, emotional engagement, micro-expressions, breathing rhythm—all of these come from a living body.

AI-generated images eliminated this intermediary; there were no actors, no bodies that had "experienced" the characters, yet Paper Cellphone still felt "real."

One possible explanation is that much of the "realism" perceived to come from the actors' bodies actually stems from the observational skills of the directors and screenwriters. When Li Ting immerses herself in the role and considers "how he will react," she draws upon her childhood memories, her observations of people, and her intuition about emotional rhythms.

These elements are passed to the model via prompts, the model generates visuals, and the visuals evoke resonance in the audience. The path has changed, but the starting point and the destination remain the same: human experience leading to human emotion .

This also explains why the partnership between the two is so effective. Yang Xuan said that Li Ting is responsible for "the imagined scenes," while he is responsible for "storytelling." When different people with different life experiences collide, what he calls "anti-plots"—details that were not in the script but were retained because they were real enough, details that AI cannot generate on its own.

In the interview, Yang Xuan mentioned the Taiwanese New Wave generation, including Edward Yang, Hou Hsiao-hsien, and Ang Lee. These directors defined "truth" as focusing on honesty on an emotional level. Hou Hsiao-hsien, in making *The Boys from Fengkuei*, used many non-professional actors, pursuing this kind of "unplanned" authenticity. He didn't want precise acting, but rather the natural reactions of people in real situations.

AI creators are inheriting this logic through different mediums. The intermediary has changed from the actor's body to the parameters of the model, but what the director injects remains unchanged: observation of people, intuition about emotions, and, as Yang Xuan repeatedly emphasizes, "sincerity."

Perfection is the enemy of creation.

Near the end of the interview, I asked a hypothetical question: If in the future AI could generate perfect works with a single click—without bugs, without any mistakes, and with every frame flawless—would you be more satisfied, or would you feel something is missing?

Li Ting's answer was straightforward: "Too much perfection isn't necessarily good."

She gave the example of the boss's character. At first, he was perfunctory with the little boy, and his excuses after chasing after him were ridiculously clumsy. But it was precisely because of this imperfection that the audience found him three-dimensional and real.

As for the bug with the telephone, Li Ting believes it's a minor flaw that doesn't detract from the overall quality. She chose that version not because of technical specifications, but because of "the characters' subtle movements and expressions, and the smoothness of the camera transitions—that's the feeling of a truly authentic performance I wanted."

"The simpler the tool, the easier it is to express yourself." Yang Xuan's answer went a step further. "You need to be clearer about what you want and what you like in order to express better things in a simpler way."

This is precisely the question that has been repeatedly discussed: as AI tools continue to iterate and technological friction decreases, what exactly is the core competitiveness of creators?

In the interview, Yang Xuan identified three key human elements: screenwriting, directing, and art direction. While "knowing how to use tools" is essential, these tools will become increasingly convenient and therefore won't constitute a competitive advantage.

It's more of a skill: knowing when to stop.

When you're emotionally invested, stop making changes; when bugs actually enhance the work, don't fix them; when leaving blank spaces is more powerful than filling them in, don't overdo it.

This kind of judgment doesn't come from models, but from human experience and intuition. The more powerful the tool, the scarcer it becomes.

Like that missing microphone, it was a technical error. But from a communication perspective, it unexpectedly accomplished something: after confirming to the audience that "this was done by AI," it made them more focused on the story itself.

This "bug" became a kind of pass, freeing viewers from the dilemma of "is this real?" because the answer was already clear. They then turned to judging a more important question: is this story good?

The answer is 40 million views and tears shed by viewers both inside and outside the film.

The most inhuman tool has produced the most human short film. Perhaps a more accurate statement is: tools have never been human. The human being is always the one using the tool.

#Welcome to follow iFanr's official WeChat account: iFanr (WeChat ID: ifanr), where more exciting content will be presented to you as soon as possible.

Wandering through the exhibition halls of the 2026 Beijing Auto Show, seeing so many similar large SUVs, a strong sense of belated realization struck me.

Unfortunately, it wasn't until today, at what was touted as the largest auto show in human history, yet utterly lacking in novelty, that we truly understood the Li Auto MEGA.

Two years ago, we included MEGA in iFanr's "2024 Tomorrow's Product List" and wrote the following in our review:

There's a strange connection between being born at the wrong time and being a product of tomorrow. The Li Auto MEGA is clearly a product of tomorrow, born at the wrong time, because a fair assessment of it lies not in the present, but in the future. Leaving aside criticism and arrogance, the MEGA, as Li Auto's first pure electric vehicle, best represents the company's product philosophy and broadens the definition of the MPV. It truly represents the height of domestic new energy vehicle product definition and completion.

When I wrote those words, it was more out of instinctive respect for a pioneering product. But two years later, as we are pushed forward by those new cars with the same face on the exhibition stand, we truly understand the weight of those words.

In 2026, automakers are sparing no effort to promote large flagship SUVs with the "9" suffix. If you put these new cars, priced at hundreds of thousands, in the same space and line them up, you will see similar smooth front ends, continuous star-ring light strips, and a lidar horn growing on the roof.

Walking into the off-road exhibition area, all you see are square metal boxes, carrying spare tire backpacks and riding all-terrain tires.

If you cover the brand logo on the front of the car, even I wouldn't dare to easily confirm the origin of these new cars.

Why are car designs becoming increasingly similar in this era of rapid technological advancement?

Some attribute this to rampant plagiarism, but this is merely a consequence, not a cause. In reality, this situation cannot be explained by a single reason. It is the result of a complex interplay of business choices, talent migration, and shifts in power dynamics.

1. The cost of trial and error is high.

Car manufacturing is a brutal game. When everyone is playing it safe, what will happen to those who dare to break the mold?

Ideal MEGA taught the entire industry an expensive lesson.

Li Xiang once shared that during the project's inception, the biggest problem the design team faced was the difficulty in reducing wind resistance. However, for a large pure electric MPV to avoid being hampered by its high-speed driving range, it is essential to reduce wind resistance.

Former Porsche designer Benjamin Baum took on the task. He and his team abandoned the "large grille plus vertical chrome bars" formula established by the Toyota Alphard, going to an extreme.

For aerodynamic reasons, unnecessary angles were removed from the body. MEGA ultimately achieved an ultra-low drag coefficient of 0.215Cd in the wind tunnel, resulting in the highly controversial teardrop-shaped front and Kambo rear.

However, the market response was like a bucket of cold water. The revolutionary design challenged the visual habits of traditional MPV users, and the flood of ridicule online caused this pioneering product to suffer greatly in the business world.

It wasn't until the subsequent launch of the i8 and i6 pure electric SUVs that Li Auto somewhat quelled the debate surrounding the exterior design, achieving a visual compromise:

It retains the MEGA's highly recognizable sloping front and star-ring lights, but at the rear, it decisively abandons the sloping coupe-like rear end and returns to the full and stable traditional SUV shape of the L series.

The front of the car weaves into the future, while the rear soothes the present.

This integration made consumers feel familiar and at ease, and it translated into solid sales figures. However, for executives at other automakers observing from the sidelines, the message this story conveyed was discouraging. Bold originality comes with high trial-and-error costs, while making minor adjustments to existing blueprints is much easier to profit from.

2. Designing "inbreeding"

Two years ago, a highly publicized dispute arose between Changan and Geely. Changan issued a lawyer's letter accusing Geely of plagiarizing its UNI-V with its debut Galaxy Light concept car, and the two sides engaged in a tense standoff on social media.

In fact, there wasn't much of a business battle involved. The reality is that Chen Zheng, who designed the Changan UNI-V, later moved to Geely's Shanghai design center and created the Galaxy Light concept car.

By tracing back the resumes of the core designers, it becomes entirely understandable that some of the "lookalikes" in the market are quite similar.

Take, for example, exterior designer Andreas Nilsson. He was the architect of Lynk & Co's first-generation design language, and was later poached by NIO, where he was deeply involved in the exterior design of NIO's second-generation platform vehicles. In 2024, he joined Changan Automobile as the head of European design.

The same applies to the interior design. Nicki Kwee, who previously served as the head of interior design within the Geely system, has worked on the interiors of Lynk & Co and Jike. He is very good at simplifying the cabin and likes to use minimalist lines and CMF (color, material, surface finish) combinations to create a sense of technology.

When he later joined NIO and took the lead in the interior design projects for the second-generation ES6, EC6, and the sub-brand Firefly, his personal style naturally followed in his new employer.

If you compare the interiors of some models from NIO, Firefly, and Geely (Lynk & Co/Jik) together, the highly similar floating center console layout and material treatment will give you a sense of familiarity, like facing "brothers from different fathers and mothers".

Such similarities in appearance or interior design are not malicious plagiarism, but rather a natural extension of the designer's personal aesthetic preferences and work habits.

When a lead designer who has spearheaded the design of a best-selling model moves to a new brand, they bring their aesthetic logic, preferred linework techniques, and even the entire team's compositional habits with them. When starting a new project, they often continue drawing in the style they are most familiar with and that has been proven successful in the market.

Several mature design languages ​​were switched back and forth between various brands by this group of top talents, and the market naturally ended up with the same face.

3. Hardware competition, with aesthetics yielding to configuration.

In the past, styling design was often a high priority in car manufacturing, and the internal mechanical layout of the vehicle sometimes even had to be compromised to achieve a smooth exterior silhouette. However, in the era of intelligent electric vehicles, with the accumulation of intelligent hardware, the designer's voice is being diluted.

Sometimes, even if designers want to stand firm, their paintbrushes have already been taken away.

The design of the Krypton 007 has caused quite a bit of controversy. This model, spearheaded by former Bentley design director Stefan Sielaff, compromised the overall proportions of the car in order to fit a large, illuminated, and interactive screen at the front.

To accommodate this thick, structurally-equipped light bar, the front of the vehicle appears relatively flat and heavy. This makes the sporty sedan appear less low-slung and more expansive in terms of visual proportions, giving it a slightly cramped feel, as if compromised to accommodate hardware.

When the industry reaches a tacit understanding that large screens, illuminated grilles, and driver assistance are synonymous with luxury, the focus of designers shifts. They no longer have ample space to meticulously refine the direction of a line or the light and shadow of a curved surface; instead, they must properly place the radar and LED panels delivered by suppliers onto the bumper or roof.

At the launch event, where the specifications were presented item by item, the elegant proportions were clearly less eye-catching than the projector headlights with millions of pixels.

Car manufacturing is moving towards "Android-like" design.

As the uniqueness of the exterior design gives way to radar and screens, the technological foundation beneath the vehicle is undergoing a more thorough unification.

In the past, the moat of luxury brands was their chassis, engines, and even genuine leather and unique stitching. Now, the supply chain has become a huge central kitchen, where new cars from different brands and with different logos not only share the same software UI and voice interaction, but even their exterior outlines are beginning to converge.

Everyone is using the same blueprints to build houses, and the market is inevitably falling into the predicament of houses becoming more and more similar.

I once interviewed a product marketing employee at a media event. I asked him directly, "Your new car, in terms of appearance, interior, and even the underlying hardware and software configuration, is very similar to another popular model on the market. In this situation, where do you think your product's competitiveness lies?"

The other party's answer was simple and straightforward:

Our price is half of itss.

This laughable answer reveals the cruelest side of homogenization. When brands cannot differentiate themselves in design and experience, a naked price war becomes the only weapon to strike at the competition.

In the first quarter of this year, domestic passenger vehicle retail sales fell by 17.5%.

The overall market is shrinking, and price wars have reached a point of no return. Against this backdrop, those highly anticipated large flagship SUVs starting with "9" have all become more conservative and hefty.

Where will this wave of homogenization sweeping across the entire industry ultimately lead the Chinese automotive industry?

Amid the hustle and bustle of the exhibition hall, many were discussing the industry reshuffling. One car company executive predicted that new brands would have anywhere from a year and a half to three years to gain market share.

Looking back at the Chinese smartphone market ten years ago, when the industry was nearing its peak, dozens of brands from other sectors simultaneously entered the market. Back then, the smartphone market was similarly homogenous, and companies were frantically piling on features.

After a brief period of euphoria came a brutal downturn, with only a handful of leading companies now dominating the market.

While the automotive industry is asset-heavy and has a longer cycle, its fundamental business principles have never changed. The current similarities are precisely the growing pains on the eve of a major industry reshuffle.

The elimination process is accelerating. Some brands are destined to fall in this cycle, forgotten by the market along with their unremarkable products. As the battle for existing market share enters its final stage, the technological advantages of the supply chain are exhausted, and simply piling on hardware will lose its effectiveness.

Convergence may be an inevitable part of the industry's explosive growth phase. However, as the market undergoes natural selection and car manufacturing returns to rationality, establishing an irreplaceable brand identity based on a unique brand philosophy is the key to a company's long-term success.

#Welcome to follow iFanr's official WeChat account: iFanr (WeChat ID: ifanr), where more exciting content will be presented to you as soon as possible.

On the evening of May 8, 2026, the afterglow of the Beijing Auto Show had not yet completely dissipated.

Born with a "capital letter" logo, the AUDI E7X has ushered in the moment of pre-sale amidst curiosity and scrutiny.

The current new energy vehicle market is noisy enough. The parameters scrolling on the big screen at the press conference, the screens getting bigger and bigger, and the computing power getting higher and higher, are gradually turning the heavy industry of car manufacturing into a competition of digital electronic products.

In this wave, how should established automakers position themselves?

The AUDI E7X delivers a very relevant answer. Relying on SAIC's intelligent platform and integrating Audi's proud chassis tuning and mechanical expertise, this new car aims to prove to the world that in an era of soaring computing power, the industrial foundation of German cars still holds an irreplaceable position.

When craftsmanship meets computing power

In the fast-paced car manufacturing cycle, pursuing efficiency is almost a survival principle for all car companies. However, Audi spent a considerable amount of time at the beginning of its press conference talking about how they "waste" time on some trivial matters.

In order to select a piece of interior leather, they follow the material selection rules of luxury handbags, and are so strict that they have to eliminate 95% of the North American A-grade top-grain leather with minor flaws.

For that 0.7 square meter piece of real wood trim on the center console, the design team went to an ecological forest at 40 degrees north latitude to find natural oak, and then used 28 processes to preserve the original growth rings and texture of the wood.

This seemingly slow approach to car manufacturing appears somewhat out of place in the rapidly evolving new energy era.

But it is precisely these physical processes that cannot be accelerated by computing power that have built the barriers that luxury brands have long maintained. When a car transforms from a simple means of transportation into a living space, the real wood grain and soft leather that you can touch with your fingertips often provide a more reassuring sense of security than a cold, impersonal screen.

All the time we seem to waste and the detours we take are for the purpose of ensuring that every layer of intelligent experience can carry the warmth and subtlety left over from the mechanical age.

This statement at the press conference perfectly encapsulates this stubbornness.

With a solid physical infrastructure in place, the subsequent development of intelligent features becomes a natural progression.

Screens and audio systems are now standard features across major brands. To differentiate themselves in a highly homogenized market, automakers need to master the nuances of subtle details. The E7X's solution is to emphasize a sense of boundaries and customization within the vehicle.

A 21.4-inch starry sky screen is installed on the roof. The panel material is QD mini LED, which is the same material used in Apple's professional displays. The peak brightness reaches 1000 nits, ensuring color performance under strong direct light.

Even more interesting attempts are reflected in the acoustic system. The E7X collaborated with Bose to create a 9.1.4.8 channel layout and pushed the speaker's DSP computing power to 2000 MIPS.

This computing power will be used for the precise segmentation of the in-car sound space. While the driver focuses on listening to navigation, the passenger can independently play immersive movie soundtracks through the ultra-near-field headrest speakers, ensuring that everyone can find their own acoustic sovereignty within the limited space of the car without disturbing each other.

While the cabin experience satisfies the senses, advanced driver assistance systems represent the real battleground for automakers. This time, Audi didn't stubbornly choose to develop cars in isolation, but decisively introduced top-tier local expertise.

As the first German luxury car in China with mass-produced Level 3 (L3) autonomous driving capabilities, the E7X's driver assistance system is a result of deep collaboration between Audi and Momenta. With the Momenta R7 reinforcement learning world model, this system demonstrated a high level of maturity in the demonstration video.

Whether it's an unmarked country road with pedestrians and vehicles sharing the road, or a narrow road U-turn that demands exceptional strategic thinking, the vehicle handles the situation with the smoothness of a seasoned driver. In the most troublesome urban parking scenarios, if there are no existing marked parking spaces, the driver can simply draw a box on the screen, allowing the vehicle to automatically calculate and park in this "created" space.

The production data from billions of kilometers of production has effectively compensated for the difficulties foreign brands face in adapting to the complex road conditions in China.

You just need to enjoy the journey, and leave the rest to Audi and Momenta.

Momenta CEO Cao Xudong, standing on the stage, made this promise. Deeply integrating authentic German mechanical quality with cutting-edge Chinese computing power may be the best path for joint venture automakers to break through in the current market environment.

The reconciliation of computing power and machinery: Car manufacturing ultimately comes down to human beings.

Intelligence and computing power make driving easier, but the bottom line of a car must ultimately return to the physical world.

In the E7X's chassis and electric powertrain system, we can still see the stubbornness of a traditional German manufacturer. All models support a 900V high-voltage platform, and with CATL's 109 kWh battery, the CLTC achieves a range of 751 kilometers. In the 10 minutes it takes to drink a cup of coffee, the vehicle can replenish its range by 429 kilometers.

Audi's renowned quattro all-wheel-drive system has evolved to version 2.0.

According to the official statement, with the assistance of air suspension, variable damping shock absorbers, and rear-wheel steering, this behemoth, which is over 5 meters long, can maintain a taut posture even when cornering. During acceleration and gentle braking, it also prevents rear passengers from experiencing the dizziness commonly associated with electric vehicles.

Beyond performance, they did something somewhat counterintuitive.

Currently, various organizations have clear standards and perspectives for crash tests, and automakers can easily strengthen their capabilities to meet these "test requirements." However, Audi, in collaboration with the China Automotive Technology and Research Center (CATARC), has developed a multi-vehicle chain random crash test.

There is no fixed collision location; only extremely high collision speeds and collision masses are set.

Once product strengths are fully developed, the next battle will be decided by market pricing.

At the end of the press conference, Stefan Poetzl, General Manager of Marketing at SAIC Audi, announced the prices of the four E7X models, ranging from RMB 289,800 to RMB 379,800.

The once dazzling array of optional extras offered by German luxury brands has been significantly reduced, replaced by a sincere approach of offering full features even at the entry level. The Pioneer base model boasts a full suite of high-value features, including advanced driver assistance systems, a 21.4-inch starry sky screen, and rear-wheel steering.

They even offered a four-seat layout in their top-of-the-line model, a rarity in its class.

With dual zero-gravity seats and a panoramic sunroof, Audi has brought an executive-level riding experience down to a price point under 380,000 yuan. By lowering its price point and offering the most generous features to meet the needs of local consumers, Audi is resolutely pursuing its electrification transformation in China.

Throughout the entire launch event, the E7X presented a wonderful sense of integration.

It is equipped with a large voice model that is rich in emotion, runs the most advanced driver assistance algorithms in China, and at the same time stubbornly polishes the texture of the original wood and fine-tunes the delicate rebound force of the shock absorbers.

I want more than just a car; I want a true life partner to help me better take care of my family and myself. And of course, the driving pleasure I love.

When Shi Pengze shared his daily life on stage, he revealed his original intention for making cars.

When the competition of parameters on the screen becomes tiring, the focus of car manufacturing must ultimately return to the human element. The coldness of machines needs to be neutralized by a soft touch, and the tedium of long-distance driving needs to be shared by reliable intelligent systems.

A line from the short video interlude at the press conference left a deep impression.

The best relationships are those where you don't have to silently shoulder the burden of the world, but instead sit side by side and hear each other's voices.

This statement perfectly encapsulates Audi's current mindset: facing the surging wave of new energy vehicles, upholding its century-old mechanical heritage and embracing cutting-edge Chinese technology have never been contradictory. The E7X acts as a bridge, connecting the glorious era of the internal combustion engine with the unknown intelligent future.

#Welcome to follow iFanr's official WeChat account: iFanr (WeChat ID: ifanr), where more exciting content will be presented to you as soon as possible.