Select Page
China’s Jidu Robo-1 Looks Like It’s From the Future. Maybe It Is

China’s Jidu Robo-1 Looks Like It’s From the Future. Maybe It Is

The collaboration with Geely could give Jidu a big boost when it comes to the notoriously tricky business of making cars at high volume and with high reliability, says Tu Le, managing director of Sino Auto Insights, an analyst firm focused on China’s automotive sector. He adds that China’s auto industry is electrifying at a faster pace than either Europe or the US because of government policies, a less entrenched gasoline-powered industry, and because such a large population allows new technologies to catch on more quickly.

JiDU Robo1 car exterior pedestrian sensor light under front headlight

Courtesy of Baidu

JiDU Robo1 car interior display

Courtesy of Baidu

The Robo-1 shows how big, innovative, and fast-moving China’s auto industry is, says Mingyu Guan, a partner at consulting firm McKinsey & Company, who focuses on the sector. Guan says that most of China’s big internet companies are developing automotive technology, in one way or another, and consumers expect an app-like experience in their vehicles. “China is like a leading beacon for the industry,” Guan says.

Baidu’s leap into automaking with Jidu is also a sign of China’s tech industry evolution. Over the past couple of years, large internet, social media, and popular app companies have faced increased regulatory scrutiny and pressure, with strict new rules around data privacy and algorithmic transparency, for instance.

The Chinese government also has signaled an intent to more tightly regulate the internet while also encouraging the development of technologies with long-term economic importance. Baidu and other firms are apparently keen to reinvent themselves by focusing on “deep tech” viewed as more valuable by the state, including technologies for electric vehicles and autonomous driving. Baidu’s most recent quarterly results, issued in May, also show that revenue from Baidu AI Cloud increased 45 percent year over year in the first quarter of 2022, while online marketing revenue shrank by 4 percent. Net losses for the period were $133 million.

Baidu has made significant investments, and received government encouragement, for autonomous driving. In November 2017, the Chinese government named Baidu one of a handful of AI “national champions” and gave the company responsibility for building an autonomous driving platform that could be used across the industry. The government’s backing also gave Baidu a leg up in working with existing automotive companies. In March the company published over 3,700 patent applications related to the technology in China. And this April, Apollo Go, Baidu’s autonomous taxi service, which operates in 10 cities in China already, received the country’s first permit for testing autonomous vehicles without a driver behind the wheel in Beijing.

Apollo also integrates with a smart-city platform that Baidu sells, and which has been adopted by 41 cities in China. This platform promises to help local authorities predict and manage congestion, road safety, and pollution using AI. Baidu CEO Robin Li touted the potential for autonomous driving to reduce road accidents, congestion, and carbon emissions in China at Baidu’s annual developer conference held in December 2021.

Jidu will no doubt be encouraged by the wider progress that China’s auto industry has made, driven in large part by the rise of electric vehicles. Chinese sales of electric vehicles jumped 169 percent in 2021 compared to a year earlier, according to data from the China Passenger Car Association, an industry organization. For 2021, electric cars accounted for 14.8 percent of Chinese car sales, compared to 4.1 percent in the US. Chinese car firms are also now exporting a growing number of EVs to Europe.

The Private Equity Firm That Grounded Paul Allen’s Dream

The Private Equity Firm That Grounded Paul Allen’s Dream

Hey, everyone. Elon now doesn’t want to buy Twitter because it can’t count its bots. You’d think an AI guy like him would let the robots speak.

Image may contain Logo Symbol Trademark Text and Label

The Plain View

Stratolaunch was based on a dream. Paul Allen, the unspeakably wealthy cofounder of Microsoft, had grown up in the thrall of space exploration, devouring books like the visionary rocketry tomes of science writer Willy Ley. In the early 2000’s, Allen funded a project known as Spaceship One, which snared the X-Prize as the first private venture to send a human into space. He later licensed the technology to Virgin Galactic, which built its own carrier vehicle to send Richard Branson into sub-orbital ecstasy. Meanwhile, Allen, frustrated with what he considered NASA’s timidity, decided to get back into the space business. He retained the legendary aircraft engineer Burt Rutan to design a giant carrier that could launch satellites and other spacecraft beyond the skies. With its twin fuselages and 385-foot wingspan, the Stratolaunch plane, later dubbed Roc, was a breathtaking spectacle in itself, doubly so because of its mission to lift its cargo to the heavens. In 2018, I trekked to the Mojave desert to see the world’s biggest airplane for myself.

But when Allen died in November 2018, after a third bout of the lymphoma that had dogged him for decades, his space dream died as well. While Stratolaunch still lives, it has no designs on crossing the Karman line. It is now an unabashed defense contractor, specializing in what the UN Office for Disarmament Affairs called “a new and destabilizing strategic weapon”: hypersonic technology that propels programmed airborne craft at speeds of Mach 5 and above.

Here’s how that happened. Upon his death, Allen’s holding company Vulcan, which included Stratolaunch, as well as sports teams and an AI think tank, fell to his sister Jodie. Apparently she had no wish to retain a space venture, offering Stratolaunch to buyers for $400 million, far less than her brother’s investment. It was unclear whether there would be any takers for the world’s biggest plane. Richard Branson, who chronically underplays Allen’s contribution to his own space venture, jokingly offered a dollar.

But one mysterious buyer emerged: Cerberus, a private equity firm named after the mythical three-headed dog who guards the gates of hell. When Vulcan made the sale in October 2019, Stratolaunch not only withheld the purchase sum, but also who bought it; reporters discovered the identity through SEC reports some months later. Maybe that was because Cerberus, run by cofounder Stephen Feinberg, has some baggage. It once tried to create a personal weapons juggernaut called Freedom Group by scooping up arms makers like Remington and Bushmaster. In 2012, Cerberus tried to divest itself of the group after a mass murderer used a Bushmaster to slaughter 20 schoolchildren and six teachers at Sandy Hook; ultimately it shifted the assets to its Remington company, which declared bankruptcy in 2018. On top of all that, Feinberg once reportedly joked that if any of his employees had their picture in the paper, “We will do more than fire that person, we will kill him.”

After buying Stratolaunch in late 2019, the private equity firm bulked up the workforce from 13 employees to more than 250 and refocused the company’s mission specifically on hypersonic vehicles. These had been considered as potential payloads during the Allen era, but they were secondary to satellite-launching and a possible manned vehicle called Black Ice. Using a carrier vehicle for hypersonic craft has its advantages; Roc can launch its rocket-propelled cargo over the ocean, where the ear-crushing sonic boom wouldn’t be so disruptive. Feinberg himself is knowledgeable about the defense establishment and served under Donald Trump as the head of the President’s Intelligence Advisory board. In December 2021, Stratolaunch won a contract from the Missile Defense Agency for a feasibility study on how the US might take countermeasures against hypersonic attacks. Stratolaunch is building its own hypersonic missiles, codenamed Talon. The first is intended for a single launch—after the test it will drop into the ocean. The second is a reusable hypersonic vehicle that will retain the key data after tests. For now, the intent is defensive, to mimic the behavior of potential attack missiles. But Stratolaunch doesn’t rule out a future role in creating offensive hypersonic weapons.

Alas, Elon Musk May Have a Point About Trump’s Twitter Ban

Alas, Elon Musk May Have a Point About Trump’s Twitter Ban

From the moment Elon Musk announced his intention to buy Twitter and impose upon it his version of free speech, speculation swirled about whether he would let Donald Trump, the ultimate Twitter scofflaw, return to the platform. Well, the suspense is over. On Tuesday, Musk confirmed what most people suspected, announcing at a Financial Times conference that he would “reverse the permanent ban” of the former president’s account. Trump, you will recall, got booted from Twitter on January 6, 2021, after his tweets during the Capitol riot were deemed to violate Twitter’s rules against glorifying violence.

As usual, the precise logic of Musk’s reasoning is hard to follow. He previously suggested that, under his ownership, Twitter would allow any content that doesn’t violate the law. But on Tuesday, he said that Twitter should still suppress tweets or temporarily suspend accounts “if they say something that is illegal or otherwise just, you know, destructive to the world.” In case that was too precise, he added, “If there are tweets that are wrong and bad, those should be either deleted or made invisible, and a suspension—a temporary suspension—is appropriate, but not a permanent ban.” 

If anything, deleting tweets that are “wrong and bad” suggests a broader, more easily abused standard of content moderation than Twitter currently deploys. (Wrong and bad according to whom?) The most likely explanation for Musk’s conflicting statements is that he’s simply making this up as he goes and has not given any serious thought to how content rules should work on the social platform that he’s trying to spend $44 billion to buy. And yet, buried in Musk’s free-speech word salad is a crouton of wisdom worth chewing on. Maybe Twitter really should rethink the use of permanent bans—not just for Trump, but for everyone.

The Trump Twitter ban has always been tough to analyze. A set of equally valid competing values point in conflicting directions. On the one hand, Twitter is a private company that can do what it wants. On the other hand, it holds an important role in American politics and public debate, such that its choices have broad consequences that bear on how democracy functions in the US. On the one hand, the public has an especially strong interest in hearing what political figures have to say; if the president has deranged or odious beliefs, that’s important information to know. On the other hand, there is something unseemly about exempting the most powerful members of society from rules that ordinary people have to abide by. Especially since rule violations by someone in Trump’s position are more dangerous than by some random Twitter user.

Getting rid of permanent bans offers one way to square these seemingly incompatible positions: In general, don’t hand out lifetime bans for average users or political figures. A permanent ban from Twitter is a harsh sentence. The platform occupies a unique place in American political life, which is precisely why Trump and other political figures are so obsessed with it. It’s where the hyper-educated “elite” who disproportionately make up the political class, especially the media, spend way too much of their time and attention. 

This is unfortunate, but it’s reality. If you want important people in media and politics to pay attention to your ideas, the best, most direct way to do that is to get into their Twitter feeds. Cutting someone off from Twitter—or from other major social platforms—can seriously constrain their ability to participate in public debate. As the Supreme Court held in 2016, “to foreclose access to social media altogether is to prevent the user from engaging in the legitimate exercise of First Amendment rights.” That was referring to an act of government, not a private enforcement decision. That distinction matters for legal purposes, but from the user perspective, the impact is the same regardless of who’s doing the banning. (Facebook at first shut Trump’s account “indefinitely” after the riot, but later agreed to the Facebook Oversight Board’s recommendation to revisit his case after a two-year suspension. YouTube has not said anything about whether or when it will let Trump back on its platform.)

Musk’s Plan to Reveal the Twitter Algorithm Won’t Solve Anything

Musk’s Plan to Reveal the Twitter Algorithm Won’t Solve Anything

“In this age of machine learning, it isn’t the algorithms, it’s the data,” says David Karger, a professor and computer scientist at MIT. Karger says Musk could improve Twitter by making the platform more open, so that others can build on top of it in new ways. “What makes Twitter important is not the algorithms,” he says. “It’s the people who are tweeting.”

A deeper picture of how Twitter works would also mean opening up more than just the handwritten algorithms. “The code is fine; the data is better; the code and data combined into a model could be best,” says Alex Engler, a fellow in governance studies at the Brookings Institution who studies AI’s impact on society. Engler adds that understanding the decisionmaking processes that Twitter’s algorithms are trained to make would also be crucial.

The machine learning models that Twitter uses are still only part of the picture, because the entire system also reacts to real-time user behavior in complex ways. If users are particularly interested in a certain news story, then related tweets will naturally get amplified. “Twitter is a socio-technical system,” says a second Twitter source. “It is responsive to human behavior.”

This fact was illustrated by research that Twitter published in December 2021 showing that right-leaning posts received more amplifications than left-leaning ones, although the dynamics behind this phenomenon were unclear.

“That’s why we audit,” says Ethan Zuckerman, a professor at the University of Massachusetts Amherst who teaches public policy, communication, and information. “Even the people who build these tools end up discovering surprising shortcomings and flaws.”

One irony of Musk’s professed motives for acquiring Twitter, Zuckerman says, is that the company has been remarkably transparent about the way its algorithm works of late. In August 2021, Twitter launched a contest that gave outside researchers access to an image-cropping algorithm that had exhibited biased behavior. The company has also been working on ways to give users greater control over the algorithms that surface content, according to those with knowledge of the work.

Releasing some Twitter code would provide greater transparency, says Damon McCoy, an associate professor at New York University who studies security and privacy of large, complex systems including social networks, but even those who built Twitter may not fully understand how it works.

A concern for Twitter’s engineering team is that, amid all this complexity, some code may be taken out of context and highlighted as a sign of bias. Revealing too much about how Twitter’s recommendation system operates might also result in security problems. Access to a recommendation system would make it easier to game the system and gain prominence. It may also be possible to exploit machine learning algorithms in ways that might be subtle and hard to detect. “Bad actors right now are probing the system and testing,” McCoy says. Access to Twitter’s models “may well help outsiders understand some of the principles used to elevate some content over others.”

On April 18, as Musk was escalating his efforts to acquire Twitter, someone with access to Twitter’s Github, where the company already releases some of its code, created a new repository called “the algorithm”—perhaps a developer’s dig at the idea that the company could easily release details of how it works. Shortly after Musk’s acquisition was announced, it disappeared.

Additional reporting by Tom Simonite.


More Great WIRED Stories

How Elon Musk Won Twitter

How Elon Musk Won Twitter

Such social media battles may be unusual when considering a takeover of a massive business, but Musk is himself unusual, says Cary Cooper, a business professor at Manchester Business School. “He’s not a traditional businessman,” he says. “He’s a man that is pretty creative and pretty innovative. He’s a unique guy and does things in a way that a normal businessperson wouldn’t do. He doesn’t play the normal games that an entrepreneur would play.”

On April 15, Twitter’s board triggered a break-glass-in-emergency financial tool: the poison pill. Also known as a limited duration shareholder rights plan, the poison pill invited shareholders to increase their investments in Twitter in order to reduce Musk’s ability to build his stake up into a controlling one. Any attempts to take his share over 15 percent would require Musk to negotiate with Twitter’s board.

Triggering the poison pill headed off the speedy hostile takeover, but Musk’s offer never left the table. On April 21, Musk outlined how he’d come up with the $44 billion in cash required to fulfill his bid. Morgan Stanley and other firms offered to back Musk’s bid, while he’d pay around $21 billion from his own estimated $263 billion fortune. The filing put meat on the bones of what had previously been a speculative offer—and indicated how seriously Musk wanted to take Twitter private.

The confirmed funding reportedly caused some of Twitter’s shareholders who were more agnostic about Musk to petition the company to hear him out. Meetings reportedly took place over the weekend, and Twitter’s board met on April 25 to recommend the deal to shareholders. It was a swift and surprising reversal. “On Friday, there was so much skepticism and cynicism, and now it almost looks like a done deal,” says Vasant Dhar, a professor of information systems at NYU Stern. Musk’s quick movements have left other potential bidders stuck playing catchup. But the deal appears to have passed the money test, at least for Twitter’s board of directors, since “the board’s fiduciary responsibility is to get the most value for shareholders,” says Galpin. “Obviously, there are questions about what he’ll do with the company if he takes control of it. He’s got to do more than just add an edit button.”

Taking the company private would allow Musk to make the changes he wants far more quickly, without answering to public markets. “I also want to make Twitter better than ever by enhancing the product with new features, making the algorithms open source to increase trust, defeating the spam bots, and authenticating all humans,” Musk wrote in Monday’s press release.

“I think he’s played it brilliantly,” says Dhar. “One could have expected the reaction we got: ‘Musk is a megalomaniac and he’s doing it for self-promotion.’ But I actually think there’s a lot more to it than that.”

It’s possible that the purchase will come under regulatory scrutiny. While there’s unlikely to be an antitrust concern, the Securities and Exchange Commission could still take issue with Musk’s disclosures along the way. “You could ask a court to enjoin the deal on the basis that he has improperly filed,” says Pritchard. “He didn’t file his initial stake on a timely basis, then he filed the wrong form because he really had the intention of influencing management the whole time,” he suggests. That, however, would require showing the harm caused by those infractions. Shareholders could lodge private lawsuits but would likely only succeed in getting more money from Musk in the deal. And the SEC is unlikely to halt the transaction because of the damage that could do to shareholders.