China's gaudy, graceless Maextro S800 is no Rolls-Royce



The Maextro S800 wants very badly to be a Rolls-Royce.

At 18 feet long, painted two-tone, lined with soft leather, backed by Huawei and built by a thousand robots in Hefei, it has the size and the price tag of ambition. What it lacks is the one thing Rolls-Royce has spent a century perfecting: restraint.

A car that can park itself is a clever feat of engineering; most Rolls-Royce owners employ a driver for that.

The Maextro comes with a 40-inch screen, roughly 40 speakers, and a party trick that lets it park itself while you film it for social media. Rolls-Royce sells the absence of gimmicks. The Maextro sells gimmicks as a feature.

Treat what follows as a cultural diagnosis. The car is just a symptom of a nation rich in cash and short on class.

Motor trend

I lived and worked in China for two years. The Maextro is the most expensive version of the kind of tacky automotive excess I saw every day on the streets of Shanghai and Chengdu.

A pearl-white BMW 7 Series gliding through traffic with a Pikachu decal the size of a dinner plate slapped on the rear door. A matte-black Porsche Cayenne with Hello Kitty stickers ringing the wheel wells. A Mercedes S-Class in a finish that violates several local optometry standards, with the owner's WeChat QR code printed on the trunk in case you wanted to add him.

People who make these choices have plenty of money. They want you to know it, immediately, from a great distance, with no possibility of misinterpretation.

The Maextro is that instinct scaled up and given a research and development budget.

Spirit of Excess

Rolls-Royce understands something the Maextro does not, which is that genuine luxury operates on the principle of subtraction. The iconic Rolls-Royce Spirit of Ecstasy hood ornament is small. The grille is dignified. Everything about the car suggests that the owner has nothing left to prove, because the proving was done by his grandfather, his great-grandfather, or some ancestor who did something morally questionable in the 1700s and was richly rewarded. Old money and new money operate on very different frequencies

China, in fairness, has had perhaps 30 years to figure out what to do with serious wealth. Desperate poverty was the default for many Chinese until relatively recently. The first generation of Chinese billionaires grew up eating cabbage in winter and now own art collections that would make a Medici blush. There is no inherited playbook for this. There is no grandfather who can pull you aside and gently suggest that the diamond-encrusted Vertu phone might be a touch much. The cultural muscle memory for restrained wealth hasn’t had time to develop, because the wealth itself is still wet behind the ears.

RELATED: Who makes the Waymos flooding American streets? China.

David Paul Morris/Bloomberg/Getty Images

From Ming to bling

So you get the Maextro: a "luxury vehicle" that confuses features with refinement, that mistakes the bill of materials for taste. Forty speakers is a number a teenager picks. A 40-inch screen is what you install when you have never considered that a car's interior might benefit from looking less like a control room. A car that can park itself is a clever feat of engineering; most Rolls-Royce owners employ a driver for that. The Maextro is engineered to impress someone standing on the sidewalk. The Rolls-Royce is engineered to impress the person sitting inside it. These are different products serving different psychologies, and only one of them is luxury.

There is something comical about watching a nation with 5,000 years of refined aesthetic standards produce a flagship sedan that resembles a karaoke lounge on wheels. This is country that gave the world Song dynasty celadon and Ming furniture so understated it still looks modern.

The classical Chinese ideal was the scholar in the bamboo grove, the brushstroke that suggests rather than declares. Somewhere between the Cultural Revolution and the iPhone, that sensibility was misplaced. What replaced it is a culture where a man worth $200 million still feels the need to wrap his Bentley in something that announces itself from a block away, because somewhere in his lizard brain, he’s still the kid whose grandmother boiled tree bark during the famine.

The Maextro will sell. It will sell to people who want a Rolls-Royce and cannot quite stomach the price and to people who want a Rolls-Royce and find the actual Rolls-Royce insufficiently exciting. It will be photographed at the entrances of exclusive nightclubs and parked outside fancy restaurants where the valets know to leave it where everyone can see it. It will do everything its buyers want a car to do.

What it won't do is fool anyone who has ridden in the real thing. Taste is built, not bought. China has the money now. The wisdom to spend it well is a generation or two behind.

OpenAI wants to make its losses public property



The only things certain in life are death, taxes, and the permanence of a government program. But what happens when a private company turns its agenda into a government program?

You cannot build a more financially secure business model than permanence. That helps explain why OpenAI is now reportedly in discussions with the Trump administration about a possible public equity stake in the company.

Unlike the dot-com bubble, whose infrastructure later supported real economic growth, rotting data centers will not leave behind comparable public value.

After all, what else is a company with $1.4 trillion in obligations and only $14 billion in revenue supposed to do?

Why was OpenAI CEO Sam Altman on Capitol Hill last week? According to the Financial Times, he was effectively selling Americans the rope to hang themselves. The plan proposed by OpenAI and other companies would reportedly create a sovereign-wealth-style fund into which AI companies would contribute equity so that the public could share in the sector’s soaring valuations.

That sounds generous until one remembers that this is still a loss-making sector built on staggering capital demands.

What is the rationale? Asked about equity stakes on Air Force One, President Trump suggested that “pieces” of AI companies could be “given to the American public” to quell growing alarm over the rapid rollout of the technology.

In other words, Americans are being asked to surrender farmland, neighborhood continuity, and the reliability of the electric grid to cloud-based, surveillance-enabling chatslop. In return, they may receive the honor of owning the losses from an insolvent business model.

The president confirmed the idea at a press conference on Wednesday, saying he would soon meet with “the top 12 or 15 executives” about “giving back something to the public.” He promised that “the public will become very rich.”

That promise should terrify everyone.

Once generative AI becomes a public project, the industry will move beyond “too big to fail.” Whatever happens to the companies or the broader sector, their success will become artificially and inextricably tied to the economy. Every government favor, subsidy, guarantee, and bailout will then be justified as necessary to protect the public’s stake.

RELATED: The AI boom is turning public meetings into crime scenes

Natalie Behring/Getty Images

Last November, OpenAI’s chief financial officer let the cat out of the bag when she said the company would need the government as a “backstop” for its business model. Sarah Friar later denied seeking a bailout. But a leaked 11-page letter from OpenAI to the Office of Science and Technology Policy urged the government to provide “grants, cost-sharing agreements, loans, or loan guarantees” to build America’s AI industrial base — all, naturally, to “compete with China.”

Fast-forward six months, and “backstop” now appears to mean a public “stake” in the company.

Everyone knows OpenAI’s generative AI model is unsustainable. It is built on unfathomably expensive capital expenditures for every token of AI usage.

Companies such as JPMorgan are reportedly finding that employees, after being pushed to use generative AI platforms such as ChatGPT and Claude, are spending more on tokens than their individual salaries. Uber’s chief technology officer said last month that the company burned through its entire 2026 budget for Claude Code and Cursor in just four months. In the irony of ironies, Microsoft itself reportedly told engineers in a major division to stop using an AI coding tool because the cost-to-utility ratio was not there.

The reality is that AI would work better through localized edge computing with low latency than through cloud-based hyperscale data centers that require unsustainable amounts of land, capital, resources, and power while causing other harms. China is producing cheap open-source AI. America is pouring concrete.

But the scale of that concrete — and all the materials, inputs, and power needed to support it — is unsustainable. Everyone knows it. Google, Amazon, Meta, Microsoft, and Oracle issued 47% more debt in the first five months of this year than they did from 2020 through 2024 combined. Total spending per capita now exceeds spending on the railroads in 1859, which at least served a clear public need that could be monetized over time.

RELATED: After fierce debate, Trump opts for federal controls in AI development

ANDREW CABALLERO-REYNOLDS/AFP/Getty Images

There is no amount of monthly household or business subscription fees that will make this investment break even. The costs will only increase because the model depends on a resource-stripping industrial footprint and GPUs that have few other useful functions and depreciate within a few years.

Unlike the dot-com bubble, whose infrastructure later supported real economic growth, rotting data centers will not leave behind comparable public value.

The tech companies, land developers, and venture capital firms understand that this is a Ponzi scheme. They are racing to take these companies public so that they can be folded into indexes, ensuring that trillions in pension funds are funneled into an unsustainable business model. Once that happens, even if a more efficient approach to AI becomes obvious, the economy and government will already be too dependent on the data center model to let it fail.

That is why these companies are also seeking federal land for their projects, a favor not extended to ordinary industries. SoftBank, the Japanese investment company trying to underwrite much of OpenAI’s speculative build-out, is reportedly pushing for a federal land project in Ohio to reduce costs. But banks are already balking at these ventures after SoftBank failed to secure a $6 billion loan for OpenAI.

Green energy taught us a simple lesson: When the only path to profitability runs through government favors, we should not start down that path.

OpenAI does not need a public stake. It needs public skepticism.

Americans should not be asked to subsidize a speculative industry, sacrifice land and power, and then call the bailout wealth creation. If AI companies cannot survive without government backstops, loan guarantees, public land, and pension-fund capture, then they are not building the future.

They are building the next permanent government program.

Republicans Launch Bills To Stop Child Predators And Foreigners From Buying Children

A new pair of bills seeks to close the loopholes that enable the surrogacy industry and its abuse of both children and immigration laws.

'Weak Link' Gavin Newsom Took $50K From Chinese EV Exec of Now-Blacklisted 'Chinese Military Company,' According to New Pentagon Classification

California governor Gavin Newsom (D.) received $50,000 in campaign contributions from a Chinese electric vehicle executive whose company the Pentagon blacklisted this week over its ties to China's military.

The post 'Weak Link' Gavin Newsom Took $50K From Chinese EV Exec of Now-Blacklisted 'Chinese Military Company,' According to New Pentagon Classification appeared first on .

Report: California’s Partnership With The CCP To Undermine American Energy May Be Illegal

China is influencing climate control policy as a form of economic warfare to silently destroy the US.

Viral robot kung-fu kicks small child, drawing mom's ire



A tourist attraction meant to captivate children ended up with one of them being beaten up.

China's flagship robots have shown to malfunction so much that it is hard to take the showcase from February as seriously as before, and now they are endangering children.

'The humanoid was described as a "Jerk clown robot."'

Months ago, Unitree, one of China's leading robotics companies, was showing off moves that seemed both futuristic and flawless. Now, the cracks have begun show in the models, along with the distinct possibility that they are being used as spy machines.

The latest hijinks from the Unitree G1 model took place at Children's Day in Xinjiang, China. The public seemingly lined up at the popular tourist spot called the Urumqi Botanical Garden, Newsflare reported, to see a kung-fu demonstration from one of the humanoid bots.

While performing a routine, a robot fitted with a blue wig took just seven seconds to kick a nearby child in the stomach.

RELATED: 'Anti-clanker': Why millions of people are cheering this android's humiliation

The routine was seemingly stopped, with the robot slowly backing off as it was likely being controlled in some manner.

The bot was described as a "jerk clown robot" by Russian outlet RT, which reported that while the child "was not seriously hurt," his mother complained that staff at the event were slow to react.

Days earlier, children were presented with similar robots in a "Waste-to-robot" event for kindergartners in Huaibei, Anhui Province, of China. Children crafted their robots out of waste for a presentation, before getting getting their hands on one of Unitree's expensive ($13,500) G1 bots.

The event was a celebration in the lead-up to Children's Day.

In late May, another G1 took a stumble seen around the world during an attempt at a Michael Jackson dance routine.

RELATED: China debuts 'scary' martial arts robots capable of backflips and weapons training

Feng Shufeng/VCG/Getty Images

While dancing to the classic song "Billie Jean," a G1 face-planted on a flight of stairs and remained motionless until it was embarrassingly dragged offstage by a staff member.

A different dance routine went off the rails in February when a street-performer bot took a tumble. The robot flailed as the likely owner tried to grab it, resulting in a swift kick in the face that reportedly left the man's nose bloody.

Axios reported on research last year that showed there was public-facing spyware installed in some of Unitree's robots, meaning anyone with the proper information could view live camera feeds without login credentials through the bots, specifically Unitree's G01 robot dogs.

Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!

Pro-China Rich Guy Neville Roy Singham Draws Growing Scrutiny Over Far-Left Activism

Spending American-made money from Shanghai as a far-left activist, Singham supports a wide range of narrative-influencing operations.

How The University Of California Became A Stooge For China’s Anti-America Warfare

The Chinese Communist Party has leveraged the University of California to shape the state's policies and prioritize China’s interests.

America desperately needs better election security



If there is a single idea that President Donald Trump holds with conviction, it is that the 2020 election was stolen.

Millions of Americans agree with him. How it was stolen, and by whom, is still being investigated six years later. That is a problem, because another national election arrives this fall, and Americans deserve an answer as to whether the way we now conduct elections can actually produce honest results.

Normal legislative remedies have failed. Congress has not passed the SAVE Act to ensure that only citizens vote, nor does it appear likely it will. It has done nothing about mass mail-in balloting or the vulnerabilities of electronic voting systems. Yet these are precisely the parts of the system that millions of Americans no longer trust — and for good reason.

The notion that the federal government has no role in federal elections is plainly wrong.

Consider what happened this past April. Treasury Secretary Scott Bessent and Federal Reserve Chairman Jerome Powell summoned the chief executives of America’s largest banks to an unannounced meeting, alarmed by a new artificial-intelligence model capable of finding and exploiting security flaws faster than any human defender could patch them.

If the men charged with protecting the nation’s financial system feel compelled to convene Wall Street on short notice over what artificial intelligence now makes possible, our election systems — built with similar computer technology but with far less security — are open to the same threat and worse.

Our electronic voting systems

For most of American history, Americans voted on paper ballots, counted by human beings, watched by other human beings. Electronic voting promised speed and accuracy. What it delivered is elections that take weeks instead of a day, accuracy that is openly in doubt, and a counting process that has lost the transparency a republic requires.

Citing proprietary software, the major vendors have become black boxes. The public is told to trust the output. Oversight is inadequate, and skepticism is the rational response.

The deeper problem is the very idea that voting and tabulation should be done electronically. The major suppliers — Election Systems & Software, Dominion Voting Systems (now Liberty Vote), and Hart InterCivic — all record and tabulate American votes on networked digital equipment running proprietary software. The vulnerability is, in part, that many of the electronic components are made in communist China. But even if all the components were made in the United States, they are not immune to a remote intrusion, a firmware exploit, or a software supply-chain attack.

The vulnerability is the architecture itself: an opaque, software-driven counting process exposed, directly or indirectly, to any determined bad actor, most especially a nation-state adversary. That is not a vulnerability at the margin. It is a structural compromise of the most sensitive function of self-government.

This is not theoretical. The People’s Liberation Army fields a cyber force approaching one million men, and American critical infrastructure is one of its principal targets. In 2019, federal officials seized a Chinese-built power transformer destined for Colorado; analysis at Sandia National Laboratory revealed what appeared to be a hardware back door enabling remote disablement.

In 2023, Microsoft identified Volt Typhoon, a Chinese campaign pre-positioning malware inside U.S. critical infrastructure to enable sabotage. To imagine that our election systems are immune to the same treatment is folly, more so now that the aforementioned use of artificial intelligence has become another weapon in the adversary’s arsenal.

Some will point to the recent Reuters account of a federal examination of Dominion machines seized from Puerto Rico, in which investigators reportedly found no Venezuelan code and only one chip sourced from China. They will conclude that the foreign-component concern has been overstated. But that misses the point entirely.

The question is not whether a particular batch of machines, examined on one occasion, contained components from a designated adversary. The question is whether a computerized voting system, however sourced and however audited, can be defended against the cyber capabilities of a nation-state intelligence service.

The honest answer is no. The same Chinese cyber force that pre-positioned malware in our power grid, water systems, and ports does not require a chip stamped in Shenzhen to reach an American voting machine. It requires only that the machine exist, be connected to a network at some point in its life, and run software that can be updated. All three conditions are met.

RELATED: John Cornyn’s defeat could be the end of the GOP establishment

Antranik Tavitian/Bloomberg/Getty Images

New evidence

Director of National Intelligence Tulsi Gabbard has declassified the Jan. 15, 2020, National Intelligence Council memorandum “Vulnerabilities in U.S. 2020 Election Infrastructure,” which judged that Russia, China, Iran, and North Korea all had “the capability to compromise U.S. election infrastructure for the 2020 presidential election.”

Senior officials briefed President Trump in February 2020. The public was never told. Subsequent declassified memos indicate that Chinese actors gained access to voter-registration databases in 12 to 18 states. Gabbard has opened a probe into allegations that intelligence officials suppressed this evidence, kept it out of the President’s Daily Brief, and hid it from Congress.

This was a serious misrepresentation by members of the Intelligence Community, especially since the Chinese Communist Party declared a "People’s War" against the United States in May 2019 in response to President Trump’s efforts to halt its theft of American intellectual property.

Communist China, which spends roughly $20 billion a year on intelligence and influence operations inside the United States, has every interest, as a matter of high government policy, in who sits in the Oval Office.

The motive could not be plainer. A second Trump term meant continued tariffs, continued enforcement against Chinese IP theft, continued pressure on Huawei, ZTE, and the Chinese semiconductor industry, and a hardening U.S. posture across the Indo-Pacific. A Biden administration meant the reversal of all of it.

Consider the numbers. Obama received 69 million votes in 2008. Clinton received 66 million in 2016. Biden received 81 million in 2020. A 15 million-vote surge for the least charismatic Democrat in living memory cannot be explained by enthusiasm.

The January 2020 assessment noted that “adversaries could also use the registration data ... to tailor other interference or influence efforts.” It is well within the realm of possibility that communist China, armed with the names on those rolls, mounted an industrial-scale effort to produce counterfeit ballots indistinguishable from genuine ones and therefore votes for Joe Biden.

Such a possibility must at least be entertained. Otherwise, one is left to ask the obvious question no one in Washington wants to ask: Why did communist China hack into those voter databases in the first place?

Can elections be secured?

Congress will not act. Blue states will not reform their mail-in practices or replace their electronic systems. Securing federal elections therefore falls to the president in his role as chief magistrate. Two executive orders are needed even if they will be challenged in court.

The first is an emergency declaration outlawing electronic voting machines in federal elections, on the grounds that any networked, software-driven counting system is inherently vulnerable to nation-state cyberattack and cannot, under current conditions, deliver an election the public can verify. Executive Order 13848 from 2018 recognized the threat of foreign interference but triggers only after the fact.

America cannot afford after-the-fact remedies.

The second would require, since the electronic voting machines would no longer be used, federal elections to be conducted on paper ballots, hand-counted by human beings observed by other human beings, with photo ID, accurate voter rolls, election-day voting, and mail-in ballots reserved for the military and the genuinely confined. The counting would be live streamed. The result would be the most transparent election in American history.

States today hold the constitutional delegation to conduct elections, and ideally, they would administer such a system themselves. Given the political divide, many will refuse. One alternative is for the federal government — preferably the National Guard, federalized and operating under each state’s adjutant general — to administer the election directly.

Critics will invoke Article I, Section 4, which empowers Congress to alter the times, places, and manner of federal elections. That route would be preferable if our political system were not broken. Others will invoke states’ rights. But states do not have rights. Citizens have natural rights, and states are obliged to defend them. When states fail to defend the most basic right of a self-governing people – the right to a fair election – the federal government has the duty to act.

RELATED: Polarization may be the cure — and the clarity — America needs

Elen11/Getty Images

The country’s critical infrastructure, which includes our election system, falls under the Department of Homeland Security. The Cybersecurity and Infrastructure Security Agency and the Election Assistance Commission sit in the executive branch. The notion that the federal government has no role in federal elections is plainly wrong.

The problem is that those agencies are not currently equipped to defend against a nation-state cyber adversary at this scale. And cybersecurity against communist China is beyond the capacity of any individual state government acting alone.

If federal authorities had actionable intelligence that a cyberattack was going to occur on America’s electronic voting systems during a federal election but did not have the ability to stop it, are they simply to stand aside and let the attack occur?

The commonsense approach would be to find a method of conducting the election that was not vulnerable to cyberattack. That is precisely why the president’s executive order is so urgently needed.

The choice at hand

As Director Gabbard’s declassifications confirm, China has gained access to the voter-registration data that defines our electorate. No election conducted on networked computers that a hostile intelligence service has the demonstrated capacity to breach, and on voter rolls that service has already breached, can deliver the legitimacy a republic requires.

Paper ballots, hand-counted, observed in the open, can. However controversial it may sound, it is the only way to ensure a fair election for the American people.

This article was originally published by RealClearPolitics and made available via RealClearWire.

America's most powerful AI superchips may be in China's hands



The U.S. built an entire export control system to keep its most powerful AI chips out of enemy hands, but a loophole may have made the system vulnerable to infiltration by China and other countries.

America's chip export rules target where a company is headquartered, not who ultimately owns it — meaning a Chinese tech giant could set up a subsidiary in Singapore or Malaysia and buy chips the parent company never could.

'The new Blackwell that just came out, it's 10 years ahead of every other chip. But no, we don't give that chip to other people.'

The Bureau of Industry and Security in the Commerce Department released new guidance Sunday, clarifying that a subsidiary of any company headquartered in a U.S. arms-embargoed nation — including China — still requires an export license to purchase advanced chips, regardless of where the subsidiary operates.

The requirement had technically been on the books since November 2023 — but the BIS acknowledged it had been receiving questions about whether it was still being enforced.

In May 2025, the bureau scrapped the Biden administration's strict AI Diffusion Rule export framework. The Trump administration called it "overly complex, overly bureaucratic," and warned that it would "stifle American innovation" and damage diplomatic relations with dozens of allied nations.

Pulling the Biden framework without a replacement in place, however, left the rules that govern who can buy these chips effectively unenforced. Furthermore, chips purchased during the loophole window do not have to be returned.

Former State Department official Chris McGuire, who helped build America's chip export framework under Biden, sounded the alarm on X Sunday, writing that "Chinese companies have been buying these chips, very likely at scale."

While the new guidance requires export licenses for subsidiaries of companies linked to U.S. arms-embargoed nations, it does not reinstate a separate safeguard: the requirement for offshore chip manufacturers to verify who is ultimately behind a purchase — a vulnerability that McGuire warned remains unaddressed.

Trump struck a deal on December 8, 2025, allowing China to purchase the H200 — a less powerful Nvidia data center chip and the company's second-best — with Nvidia paying 25% of those sales back to the U.S. government. Trump announced the deal on Truth Social, writing that Chinese President Xi Jinping "responded positively."

However, it appears China "chose not to" approve the purchases.

RELATED: Trump goes to China with THIS surprise guest and an entourage of business leaders

Brendan Smialowski/AFP/Getty Images

While China was passing on the H200, a scaled-down export variant, Nvidia's Blackwell chips — which defense analysts warn could serve as the foundation for next-generation autonomous weapons systems — may have been flowing freely through the back door.

The Blackwell chip was never supposed to reach any entities linked to U.S. arms-embargoed countries. Trump made that explicit while speaking to reporters aboard Air Force One in November 2025: "The new Blackwell that just came out, it's 10 years ahead of every other chip. But no, we don't give that chip to other people."

Al Jazeera reported that Nvidia said the company had been operating according to the clarified rules, claiming its "sales and vetting process is correct." Nvidia also claimed China "has more than enough domestic chips for all of its military applications," raising questions about the Chinese military actively seeking Nvidia chips in the first place.

Nvidia did not respond to a request for comment from Blaze News.

Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!