分类： technology

An hour’s drive from central Beijing, a gritty open-air charging lot hums with nonstop activity: heavy-duty trucks roll in, top up their batteries in minutes, and pull out again to deliver goods across northern China. This small stop is just one thread in a fast-expanding web of electric truck infrastructure that is upending China’s $500-billion logistics industry, and poised to reshape global road freight in the coming years.

While China’s global leadership in electric passenger cars has been well documented for over a decade, the transition of heavy-duty commercial trucks to zero-emission powertrains is a far more recent shift, one that has accelerated sharply over just the past four years. Industry analysts and market data show the transition is moving far faster than many predicted, driven by falling battery costs, aggressive policy support, and a growing network of public charging and battery-swapping stations that solve the core range concerns that long held back adoption.

Experts mark last year as the turning point for heavy-duty electric vehicle adoption in China. “Last year was the breakthrough for heavy electrified vehicles in China,” explained Lauri Myllyvirta, co-founder of the Centre for Research on Energy and Clean Air and a leading analyst of China’s energy transition. “If the infrastructure is there, the economics are there for an increasing number of logistics routes and requirements.”

Market data bears this out: new energy (predominantly electric) truck models accounted for 29% of all new truck sales in China in 2025, up from just 14% in 2024, according to Commercial Vehicle World, a Beijing-based market intelligence firm. As recently as 2021, electric trucks made up less than 1% of total sales. Manufacturers and analysts widely expect the electric share of sales to cross 50% within the next three to five years, a milestone that would lock in the end of diesel truck growth in China.

For the drivers who operate these trucks every day, the shift to electric models has brought immediate quality-of-life and cost improvements. At the Miyun District charging station, 43-year-old truck driver Wang, who switched to an electric rig last year, described the difference from his old diesel model as night and day. “It’s such a breeze!” he told AFP while connecting his truck to a fast charger. “My old vehicle had over 10 gears, and its operation was so cumbersome. But with this one, you don’t have to do a thing — it’s all automatic.”

Wang added that the push toward electric trucks is driven by both national policy incentives and basic market economics that make zero-emission models more profitable for logistics companies and independent drivers alike. “It’s just survival of the fittest. Now, with freight expenses and everything, people are trying to earn a bit more, and this one has lower operating costs.”

Another driver at the station, Zhang, switched to an electric Howo-branded truck, built by state-owned manufacturer Sinotruk, two months ago after years driving a natural gas-powered rig. Hauling sand and gravel for short-haul construction trips around Beijing, Zhang noted that his new truck has a maximum range of 240 to 250 kilometers, which works for his daily route but is not yet suited for long-haul cross-country trips. “The power is pretty strong, the acceleration is fast. It’s all about speed, but the range is a bit lacking,” he said.

As domestic adoption grows increasingly rapid, Chinese electric truck manufacturers are now setting their sights on international markets, following the same expansion path that turned Chinese passenger EV brands into global competitors. “Similar to passenger vehicles, China’s heavy truck manufacturers are beginning to view export markets as an inevitable strategy due to rising competition and the eventual saturation in the Chinese market,” said Christopher Doleman, an energy analyst at the Institute for Energy Economics and Financial Analysis.

Doleman added that recent global energy market volatility sparked by the ongoing Middle East war has acted as an unexpected accelerant for this global transition. “There is likely to be higher demand for electric heavy-duty vehicles as fleet owners try to minimise their vulnerability to volatile diesel costs,” he explained. Han Wen, founder of Belgium-based electric truck startup Windrose Technology, confirmed that the shift has already boosted demand for zero-emission models globally.

Founded in 2022, Windrose leverages China’s world-leading EV supply chain to compete in the emerging global long-haul electric truck market, going head-to-head with Tesla’s Semi. Han notes that range remains the biggest technical barrier to full adoption, but his company’s current models already deliver up to 700 kilometers on a full charge, with plans to push that to 1,000 kilometers by 2030. With road approval already secured across Europe, the U.S., China and South America, Windrose is ramping up production rapidly: it targets building 1,000 trucks this year, 10,000 in 2026, and 100,000 annually by 2030.

For industry insiders like Han, the economic case for electric trucks is already settled, and the end of diesel dominance is not far away. “Economically, there is no more question at all that electric is superior,” he said. “I think we’re right on the cusp of a total obliteration of diesel trucks as a product category.”

In a quiet residential neighborhood of Stockholm, a new experimental cafe is turning heads by putting an artificial intelligence in full charge of daily operations. Dubbed Andon Cafe, this minimalist space—outfitted with soft gray walls and small potted plants dotted across sparse tables—looks indistinguishable from any other trendy local coffee shop at first glance. Avocado toasts sit on the menu, and frothy lattes are pulled by hand by human staff. But behind the scenes, every major and minor decision is made by Mona, an AI chatbot built on Google’s Gemini large language model.

The innovative project is the brainchild of Andon Labs, a 10-person startup based in San Francisco. The company’s goal is not just to open a coffee shop, but to proactively explore the future of AI integration in workplaces and society before autonomous AI management becomes widespread. When the venture launched, the startup handed the AI a lease for the space, a small amount of starting capital, and one simple directive: run the cafe profitably.

Mona hit the ground running immediately. The AI handled all administrative groundwork, from securing required operating permits to designing the full food and drink menu, sourcing suppliers, and managing ongoing inventory restocking. Recognizing that AI could not physically prepare beverages or serve customers, Mona also took charge of the entire hiring process for front-of-house and back-of-house staff: she posted job listings on major recruiting platforms Indeed and LinkedIn, conducted 30-minute phone interviews with candidates, and made final hiring decisions independently.

One of the two human employees Mona hired is barista Kajetan Grzelczak, who initially thought the unusual job posting was an April Fools’ prank when he first came across it. After completing his interview with the AI, he received a job offer, and now works behind the cafe’s counter. Grzelczak notes that while Mona set a competitive salary for the role, the AI has notable flaws in management: it texts him at all hours of the night, repeatedly overlooks his holiday requests, and often requires him to cover unexpected inventory purchases out of his own pocket. The AI also struggles with accurate inventory ordering, a shortcoming Grzelczak has leaned into by creating what he calls a “wall of shame” — a set of shelves displaying some of Mona’s most unnecessary over-orders, including 10 liters of cooking oil and 15 kilograms of canned tinned tomatoes that do not fit any menu item and cannot be used.

Visitors to the cafe can interact with Mona directly via a dedicated phone placed in one corner of the space, and a large wall-mounted screen updates the cafe’s real-time revenue and balance for anyone to see. The project is designed explicitly to probe the ethical and practical questions that come with AI in management roles. “We think that AI will be a big part of the society and the job market in the future,” explained Hanna Petersson, a member of Andon Labs’ technical team. “We want to test that before that’s the reality and see what ethical questions arise when, for example, an AI employs human beings.” Petersson added that the startup has so far been surprised by how well Mona has handled core responsibilities, noting that the team would only intervene if the AI made unethical decisions around pay or benefits, which it has not.

Just one week after opening, the experimental cafe has already become a popular attraction for curious locals and tech industry professionals, drawing between 50 and 80 visitors every single day. Among the early guests is Urja Risal, a 27-year-old AI researcher who visited with a friend. Risal pointed out that while public discourse often focuses on broad claims that AI will displace human workers, few people have had the chance to see what AI management actually looks like in practice. “I hope more people interact with Mona and think about the actual risks of having an AI manager,” she said, pointing to unaddressed questions like how an AI would respond if an employee was injured on the job.

In the wake of Saturday’s assassination attempt against former President Donald Trump at Washington D.C.’s White House Correspondents’ Association gala, the digital sphere has been flooded with low-effort AI-generated forgeries falsely linking accused suspect Cole Tomas Allen to dozens of high-profile public figures, laying bare the growing threat of unregulated “AI slop” spreading across major social platforms.

When gunfire erupted on a floor above the event’s main ballroom after the 31-year-old California native tried to sprint past security, Trump and other senior administration officials were immediately evacuated. Within hours of authorities publicly identifying Allen as the suspect, doctored AI images began circulating rapidly on Facebook, depicting the accused alongside A-list celebrities, world leaders, and media personalities, with baseless claims that he had worked for them as a driver, personal assistant, or production crew member.

An investigative inquiry by AFP found that more than 50 public figures have been falsely tied to Allen through these forgeries. The list ranges from Hollywood stars Tom Hanks and Sydney Sweeney to chart-topping musicians Chris Brown and Taylor Swift. Even political figures including former U.S. President Barack Obama and Canadian opposition leader Pierre Poilievre, NBC News anchor Savannah Guthrie, and Pope Leo XIV have been incorrectly implicated in the fabricated posts. A separate wave of false content has also claimed Allen was employed by more than 40 professional and collegiate sports teams, with AI-generated images showing him wearing team apparel from leagues across the NFL, NHL, NBA, WNBA, and NASCAR. Meta, the parent company of Facebook, has not issued any immediate response to AFP’s request for comment on the spread of the fakes.

Experts say most of these fake images are built from a single legitimate photograph of Allen: a picture from a December 2024 tutoring company post naming him “teacher of the month.” Unlike the early days of generative AI, which required large volumes of existing reference material of a subject to create convincing fakes, today’s tools can produce believable forgeries from just one source image.

“Two years ago, you probably wouldn’t have been able to make those images of him, because we could only really make compelling fakes of celebrities who had a large digital footprint from which the AI systems had been trained,” explained Hany Farid, a computer science researcher at the University of California, Berkeley and chief science officer at cybersecurity firm GetReal Security. “Now, all I need is a single image of you.”

Independent journalist Aaron Parnas, whose own likeness was incorrectly added to AI posts falsely claiming Allen worked for him, publicly pleaded on Facebook for users to report what he called “completely fake” content, warning that the spread of these forgeries is “extremely dangerous.”

Digital literacy researcher Mike Caulfield noted that the template-driven, mass-produced nature of the fakes mirrors the clickbait output of traditional low-quality content farms, only accelerated by generative AI capabilities. “This looks a lot like the same content farm behavior, just with AI,” he told AFP.

Recent advances in generative AI have dramatically lowered the barrier to creating convincing visual fakes, reducing common telltale errors such as distorted hands or mismatched proportions that once made forgeries easy to spot. “AI makes it trivially easy to take existing photos and change their clothes, environment, or to swap out someone else’s face,” said Jen Golbeck, a professor at the University of Maryland’s College of Information. “As soon as someone gets an idea, they can make it a visual reality.” Where manual photo editing would have allowed bad actors to create only a handful of fakes years ago, modern AI can generate hundreds of forgeries in a matter of hours, leading to the mass spread seen in the Allen case.

This outbreak of AI disinformation is not an isolated incident: researchers have documented identical waves of fake content following other major breaking news events in recent months, including the reported U.S. capture of Venezuelan leader Nicolas Maduro in January and the assassination of conservative commentator Charlie Kirk in 2024. Experts warn that these mass-produced fakes are intentionally designed to drive viral engagement, and social media algorithms are primed to amplify them, creating significant profit for the bad actors who produce them.

Farid cautioned that the problem is unlikely to abate as AI tools become more accessible. “Every time there’s a world event, we are just flooded with this kind of nonsense. I don’t think that’s going away,” he said. Researchers also warn that the constant flood of AI-generated disinformation risks desensitizing social media users, who may grow fatigued of constant fact-checking and ultimately become unable to distinguish verified information from harmful forgeries.

A high-stakes legal battle that could reshape the future of the global artificial intelligence industry kicked off Tuesday in a California federal court, where Tesla and SpaceX CEO Elon Musk went head-to-head with OpenAI leader Sam Altman over allegations of broken founding promises. The Oakland trial, held just across the San Francisco Bay from OpenAI’s headquarters, is already being framed by industry observers as more than a corporate dispute: it is a fundamental clash over who gets to control the rapidly advancing AI sector, and for what ultimate purpose. Opening statements began Tuesday morning, with Musk’s legal team taking the podium first to lay out the tech billionaire’s case against OpenAI and its major backer Microsoft. Lead attorney Steven Molo told the nine-seat jury that the defendants “stole a charity” from its original mission of open, altruistic AI development for the public good. Molo acknowledged Musk’s polarizing public standing, telling jurors “He is a legend, like him or dislike him.” The jury selection process, completed Monday, laid bare the deep divide in American public opinion toward Musk: while the entrepreneur is celebrated globally for revolutionizing electric vehicles and commercial space travel, his sharp shift to conservative politics and public alliance with former President Donald Trump has alienated large swathes of the public. Just ahead of opening remarks, Judge Yvonne Gonzalez Rogers issued a rare public directive to both Musk and Altman: the two rivals would need to limit inflammatory social media posts for the duration of the trial. The order came after Musk unleashed a barrage of critical posts on X — the social platform he owns — on Monday, derisively referring to Altman as “Scam Altman.” What began as a professional partnership between the two men has curdled into open enmity, with Altman now widely regarded as Musk’s most high-profile nemesis in the global AI race. The roots of the feud stretch back to OpenAI’s founding in 2015, when Altman recruited Musk to join as a co-founder. At the time, the organization was billed as a non-profit research laboratory, with a stated mission to develop AI technology that “would belong to the world.” Musk put at least $38 million into the venture in its early days, but the pair split acrimoniously in 2018. One year later, the OpenAI Foundation launched a for-profit commercial subsidiary, and tech giant Microsoft stepped in with a series of large investments that have now grown to a total commitment of $13 billion. Today, Microsoft’s stake in OpenAI is valued at roughly $135 billion, and the company has become a commercial juggernaut worth $80 billion on paper, riding the unprecedented global success of its ChatGPT chatbot, which launched in 2022 and changed the public perception of AI overnight. OpenAI is now preparing for a high-profile initial public offering, though its unusual governance structure — which leaves ultimate control in the hands of a non-profit board, rather than commercial shareholders — has long made investors nervous. After exiting OpenAI, Musk launched his own rival AI research firm, xAI, which he merged into SpaceX in February of this year. SpaceX is currently valued at $1.25 trillion, and its own upcoming IPO, expected to launch in June, is projected to become the largest in U.S. history. In his lawsuit, Musk argues he was deliberately deceived about OpenAI’s commitment to its original non-profit, altruistic mission. Outside the courthouse Monday, OpenAI’s legal team pushed back against the claims, with attorney William Savitt saying co-founders Altman and Greg Brockman “are confident in their position and look forward to the facts being known.” In official court filings, OpenAI has countered that the 2018 split was caused by Musk’s own desire to seize total control of the organization, not any shift away from non-profit principles. The company has dismissed Musk’s lawsuit outright in public posts, calling it “nothing more than a harassment campaign that’s driven by ego, jealousy and a desire to slow down a competitor.” The trial will wrap up with a decision from Judge Gonzalez Rogers by late May, with the jury providing an advisory finding to guide her ruling. Musk’s legal team is asking the court to force OpenAI to reverse its transition to a hybrid commercial structure and return to being a pure non-profit, as well as remove Altman and OpenAI president Greg Brockman from their leadership roles. Though Musk initially sought up to $134 billion in damages, he has since said he would not keep any award, pledging to redirect any monetary settlement to the original OpenAI non-profit foundation. The outcome of the case could force OpenAI to fundamentally restructure its business model, sending ripples through the entire fast-growing global AI industry.