A Chinese Wikipedia editor spent years writing fake Russian medieval history

Chinese Wikipedia had a robust collection of detailed and authoritative articles on medieval Russia, thanks to a user called Zhemao who claimed to be the daughter of a diplomat assigned in the country. Zhemao wrote 206 articles for the website since 2019, the longest of which, according to Vice World News, is almost as long as The Great Gatsby. It detailed Tartar uprisings in 17th century Russia and was supported by a map of the country during that era. In another article, the user shared rare images of ancient Russian coins that were purportedly obtained from archeologists. The articles she contributed were so well-written and well-regarded, until it was revealed that she’d pulled off one of the largest hoaxes ever seen on the platform.

Chinese novelist Yifan was the one who brought the hoax to light in a post on a website similar to Quora. Yifan stumbled upon one of her articles describing a silver mine that provided a source of wealth for Russia in the 14th and 15th centuries while researching for a new book. The article was reportedly so detailed, it included information on the composition of the soil, the structure of the mine and the refining processes used on the silver. But when Yifan tried to fact check Zhemao’s references with Russian speakers, it was revealed that the pages or the versions of the books she cited didn’t even exist. 

A group of volunteer editors combed through her work as a response and found that her citations didn’t add up or that she fabricated information from legitimate sources that were too obscure to be fact-checked by casual users. As a crowd—sourced online encyclopedia, Wikipedia trusts its contributors to self-regulate. In an article about its reliability, Wikipedia said it maintains an inclusion threshold of “verifiability, not truth.” 

A volunteer editor who’s been helping clean up articles Zhemao contributed to told Vice News that they only typically check articles for blatant plagiarism and to ensure that they’re properly cited. That is why vandalism is a common occurrence on the website and why its reputation as a legitimate source of knowledge is frequently challenged. Chinese Wikipedian John Yip told Vice that “Zhemao single-handedly invented a new way to undermine Wikipedia.” It’s worth noting, however, that she’s far from the first person to pretend to be an expert on the website. Back in 2007, an editor who claimed to be a university professor was revealed to be a 24-year-old from Kentucky who had no higher-education credentials.

Zhemao, in a post on her profile, has admitted to making up her whole identity and to fabricating information. She came clean that she’s not based in Russia and that her husband is not Russian but Chinese. She also doesn’t have a doctoral degree in world history from the Moscow State University like she claimed, but is instead a housewife with a high school diploma. Vice said, based on her post, that she got frustrated about not being able to understand articles in Russian and in English. She apparently used online translators to understand articles available online and then used her imagination to fill in the gaps.

Why she didn’t just write a novel set in medieval Russia — and it could’ve been a hit, based on how Yifan and her fellow editors praised her contributions for being thorough and well-written — is not quite clear. Zhemao and her sock puppets had been banned permanently from the website, though, so she might end up looking for a new outlet for her writing.

Kevin Spacey Pleads Not Guilty To Sexual Assault Charges

The Oscar winner was charged with sexually assaulting three men between 2005 and 2013.

Jury convicts ex-CIA engineer for leaking the agency's hacking toolset

Joshua Schulte, the former CIA engineer arrested for what’s being called the biggest theft of classified information in the agency’s history, has been convicted by a federal jury. Schulte was arrested in relation to the large cache of documents that Wikileaks had published throughout 2017. That string of CIA leaks known as “Vault 7” contained information on the tools and techniques the agency used to hack into iPhones and Android phones for overseas spying. It also had details on how the CIA broke into computers and how it turned smart TVs into listening devices. A federal jury has found Schulte guilty on nine counts, including illegally gathering national defense information and then transmitting it.

According to The New York Times, Schulte was arrested after investigators traced the leaks to him. The former CIA engineer worked with a team in a secret building protected by armed guards to create tools, like malware, that were used to target the devices of suspected terrorists. In 2018, he was formally charged with 13 counts that included theft of classified information, obstruction of justice, as well as possessing and sending images and videos with child pornography. He’s still awaiting trial on charges of possessing child pornography, which he allegedly downloaded from 2009 until March 2017. 

Schulte’s original trial back in 2020 was declared a mistrial after jurors couldn’t come to an agreement regarding some of hist most serious charges, illegally gathering and transmitting national defense information included. After that event, the former CIA engineer had decided to represent himself. As part of his closing arguments, he told the jurors that the CIA and the FBI made him a scapegoat for their embarrassing failure, repeating what his side had been saying from the time he was arrested.

While the judge, AP said, was impressed with his closing arguments, they weren’t enough to get the jury on his side. In court, he argued that the government’s case is full of holes and that he didn’t even have motive to leak the CIA’s hacking tools. Prosecutors, however, accused him of being a disgruntled employee who felt that he was disrespected when the agency ignored his complaints about his work environment. As retaliation, he allegedly tried “to burn [the CIA] to the ground.” US Attorney Damian Williams said his actions rendered the “most valuable intelligence-gathering cyber tools used to battle terrorist organizations and other malign influences around the globe” essentially useless. Williams also accused Schulte of trying to leak more classified materials against the government while he was behind bars. 

Schulte will have to face the court again to face charges related to possession of child pornography before a sentencing date can be set. The nine counts he was convicted of, however, are enough to keep him in prison for up to 80 years.

Meta's 'Make-A-Scene' AI blends human and computer imagination into algorithmic art

Text-to-image generation is the hot algorithmic process right now, with OpenAI’s Craiyon (formerly DALL-E mini) and Google’s Imagen AIs unleashing tidal waves of wonderfully weird procedurally generated art synthesized from human and computer imaginations. On Tuesday, Meta revealed that it too has developed an AI image generation engine, one that it hopes will help to build immersive worlds in the Metaverse and create high digital art.

A lot of work into creating an image based on just the phrase, “there’s a horse in the hospital,” when using a generation AI. First the phrase itself is fed through a transformer model, a neural network that parses the words of the sentence and develops a contextual understanding of their relationship to one another. Once it gets the gist of what the user is describing, the AI will synthesize a new image using a set of GANs (generative adversarial networks).

Thanks to efforts in recent years to train ML models on increasingly expandisve, high-definition image sets with well-curated text descriptions, today’s state-of-the-art AIs can create photorealistic images of most whatever nonsense you feed them. The specific creation process differs between AIs.

a bunch of nonsense
Meta AI

For example, Google’s Imagen uses a Diffusion model, “which learns to convert a pattern of random dots to images,” per a June Keyword blog. “These images first start as low resolution and then progressively increase in resolution.” Google’s Parti AI, on the other hand, “first converts a collection of images into a sequence of code entries, similar to puzzle pieces. A given text prompt is then translated into these code entries and a new image is created.”

While these systems can create most anything described to them, the user doesn’t have any control over the specific aspects of the output image. “To realize AI’s potential to push creative expression forward,” Meta CEO Mark Zuckerberg stated in Tuesday’s blog, “people should be able to shape and control the content a system generates.”

The company’s “exploratory AI research concept,” dubbed Make-A-Scene, does just that by incorporating user-created sketches to its text-based image generation, outputting a 2,048 x 2,048-pixel image. This combination allows the user to not just describe what they want in the image but also dictate the image’s overall composition as well. “It demonstrates how people can use both text and simple drawings to convey their vision with greater specificity, using a variety of elements, forms, arrangements, depth, compositions, and structures,” Zuckerberg said.

In testing, a panel of human evaluators overwhelmingly chose the text-and-sketch image over the text-only image as better aligned with the original sketch (99.54 percent of the time) and better aligned with the original text description 66 percent of the time. To further develop the technology, Meta has shared its Make-A-Scene demo with prominent AI artists including Sofia Crespo, Scott Eaton, Alexander Reben, and Refik Anadol, who will use the system and provide feedback. There’s no word on when the AI will be made available to the public.

The Morning After: Hasbro can 3D-print your face onto your favorite action figure

The Hasbro Selfie Series is a collaboration between the toy maker and 3D printing specialists Formlabs, can customize an action figure with your own face.

TMA
Hasbro

You can scan your face with a smartphone and get a custom-made, look-a-like action figure with opposable limbs: AKA the dream. At launch, options include Star Wars X-Wing pilots, Ghostbusters, Power Rangers, Snake Eyes from GI Joe and a few more. The idea seems to have come from a comic-con competition that offered custom action figures for five winners.

You’ll need to download Hasbro Pulse, the company’s dedicated mobile app, and wait for the scan feature to open up closer to the expected ship date in the Fall. The six-inch, “collector-grade” figure will cost $60 (plus taxes). Unfortunately, the offering is only available to customers in the US. I wonder if the majority of figures ordered will have childrens’ faces, or their parents’?

– Mat Smith

The biggest stories you might have missed

Magic Leap 2 will go on sale in the US on September 30th

The enterprise headset’s price will start at $3,299.

Magic Leap’s second-generation mixed-reality headset finally has a concrete commercial release date. Magic Leap 2 first made its way to a limited number of users last year as part of the company’s early adopter program. Starting on September 30th, though, anybody who wants (and can afford) to get the headset in the US can buy one. It’ll launch in three editions, the cheapest of which is the Base headset meant for professionals and developers who just want access to the augmented reality platform. The edition can be used for full commercial deployments and production environments and will cost $3,299 with a one-year warranty.

Continue reading.

Google files a lawsuit that could kick Tinder out of the Play Store

It’s counter-suing Match Group over in-app payments.

Earlier this year, Match sued Google alleging antitrust violations over a decision requiring all Android developers to process “digital goods and services” payments through the Play Store billing system.

Following the initial lawsuit in May, Google and Match reached a temporary agreement allowing Match to remain on the Play Store and use its own payments system. However, Google parent Alphabet claims that Match Group now wants to pay “nothing at all” to Google, including its 15 to 30 percent Play Store fees.

Continue reading.

Preview: Apple’s watchOS 9 beta

Upgrades for workout and sleep tracking.

TMA
Engadget

With watchOS 9, the company is bringing a robust slate of Workout updates, alongside new watch faces, redesigned apps and the ability to detect sleep zones. Thanks to the public beta, we’ve been able to take a closer look. We’re fascinated by the addition of cardio zones to the workout features. It might even be worth the risk involved in running beta software.

Continue reading.

PS Plus Extra and Premium games for July include ‘Stray’ and ‘Final Fantasy VII Remake Intergrade’

You’ll get access to ‘Marvel’s Avengers’ and a bunch more Assassin’s Creed games too.

Sony has announced the first new batch of games for its higher PlayStation Plus tiers since it revamped the service. Extra and Premium subscribers will have access to Stray, a cyberpunk adventure game in which you play as a cat with a drone companion. It’s the first time that a third-party title will be available on a higher PS Plus tier on its release date — something that happens on Microsoft’s Game Pass a lot.

Continue reading.

Sega’s Genesis Mini 2 hits North America on October 27th

With six-button controller innovation.

TMA
Sega

Sega is bringing the Genesis Mini 2 to North America on October 27th. The company previously said its latest retro console will go on sale in Japan on the same day (though it’s correctly called the Mega Drive Mini 2 there). It will come with more than 50 games, including Sega CD titles. At the time of writing, the Amazon listing shows that the console will ship from Japan to the US for $103.80 plus a $22 delivery fee.

Continue reading.

How Often Is Long Covid Happening? The Answer Isn't So Easy to Find

Three years in, the impacts of the covid-19 pandemic continue to reverberate around the world. The widescale, if unequal, distribution of vaccines has surely saved millions of lives and lessened people’s misery since the beginning of 2021. But thousands of people worldwide are still dying, and millions are getting…

Read more…

The Weirdest Things IT Workers Have Found On Company Laptops

From secretly snapped feet pics to love letters to colleagues, tech workers find some pretty strange stuff on company computers.

Uber sued by more than 500 women over sexual assault and kidnapping claims

Uber is facing a lawsuit filed by more than 500 women who allege they were assaulted by drivers, CNBC has reported. The complaint states that “women passengers in multiple states were kidnapped, sexually assaulted, sexually battered, raped, falsely imprisoned, stalked, harassed, or otherwise attacked,” by Uber drivers. The San Francisco law firm that filed the suit said it has about 550 clients with at least another 150 claims being investigated. 

Earlier this month, Uber released its second safety report showing that sexual assault reports in the five most severe categories fell 38 percent from 5,981 in 2017 and 2018 to 3,824 for the years 2019 and 2020. However, that may be correlated with the COVID-19 pandemic which saw a severe drop in ridership from 2020-2021. “We’re constantly innovating and investing in the safety of our platform,” Uber chief legal officer Tony West wrote in the report.

However, the law firm said that safety is not the company’s highest priority. “Uber’s whole business model is predicated on giving people a safe ride home, but rider safety was never their concern – growth was, at the expense of their passengers’ safety,” said Slater Slater Schulman LLP founding partner Adam Slater. “While the company has acknowledged this crisis of sexual assault in recent years, its actual response has been slow and inadequate, with horrific consequences.”

The law firm criticized Uber for lax policies related to driver background checks and enforcement. It noted that Uber has “opted to hire drivers without fingerprinting them or running their information through FBI databases… [and] has a longstanding policy that it will not report any criminal activity – even assaults and rape – to law-enforcement authorities.” 

Uber has yet to respond to the lawsuit, but Engadget has reached out for comment. An Uber spokesperson told Fox Business that it can’t comment on pending litigation, but that the company “takes reports of this nature very seriously and has worked closely with advocates to develop a survivor-centric approach to handling such cases when they arise.” 

Uber has a history of settlements and complaints related to passenger and driver safety. In 2016, The Guardian reported that Uber had paid out $161.9 million in safety-related lawsuits since 2009. In 2017, it faced a class-action lawsuit accusing it of “giving perpetrators of sexual assault, sexual harassment and physical violence access to thousands of ‘vulnerable victims’ nationwide.” And in 2019, the company was sued for $10 million by a woman who was sexual assaulted by an Uber driver, saying the company put her in harm’s way. 

Celsius Files for Bankruptcy as Users Worry Their Money is Gone Forever

Celsius filed for Chapter 11 bankruptcy protection late Wednesday, leaving users wondering if they’ll ever again see the money they put into the crypto platform. Unfortunately for the average user, they’ll have to wait in line with the company’s other creditors, something that was made explicit in the platform’s…

Read more…

Man Accused Of Killing Friend Claims It Was To Stop Him From Summoning Sasquatch

He reportedly feared that Bigfoot was coming to eat him.