You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.
A selection of 13 links for you. Thank goodness you read this bit. I’m @charlesarthur on Twitter. Observations and links welcome.
Last Friday, a portrait produced by artificial intelligence was hanging at Christie’s New York opposite an Andy Warhol print and beside a bronze work by Roy Lichtenstein. On Thursday, it sold for well over double the price realized by both those pieces combined.
“Edmond de Belamy, from La Famille de Belamy” sold for $432,500 including fees, over 40 times Christie’s initial estimate of $7,000-$10,000. The buyer was an anonymous phone bidder.
The portrait, by the French art collective Obvious, was marketed by Christie’s as the first portrait generated by an algorithm to come up for auction. It was inspired by a sale earlier this year, in which the French collector Nicolas Laugero Lasserre bought a portrait directly from the collective for about 10,000 euros, or about $11,400.
Mr. Rubin often berated subordinates as stupid or incompetent, they said. Google did little to curb that behavior. It took action only when security staff found bondage sex videos on Mr. Rubin’s work computer, said three former and current Google executives briefed on the incident. That year, the company docked his bonus, they said.
Mr. Singer, the spokesman for Mr. Rubin, said the executive “is known to be transparent and forthcoming with his feedback.” He said Mr. Rubin never called anyone incompetent.
Mr. Rubin, 55, who met his wife at Google, also dated other women at the company while married, said four people who worked with him. In 2011, he had a consensual relationship with a woman on the Android team who did not report to him, they said. They said Google’s human resources department was not informed, despite rules requiring disclosure when managers date someone who directly or indirectly reports to them.
In a civil suit filed this month by Mr. Rubin’s ex-wife, Rie Rubin, she claimed he had multiple “ownership relationships” with other women during their marriage, paying hundreds of thousands of dollars to them. The couple were divorced in August.
The suit included a screenshot of an August 2015 email Mr. Rubin sent to one woman. “You will be happy being taken care of,” he wrote. “Being owned is kinda like you are my property, and I can loan you to other people.”
These two journalists have been researching this story for about a year, they say; others have also been trying to pull it together.
But also, it’s not only about Rubin. It’s other men who were in senior positions, had credible accusations made against them, and then were forgiven or given big payoffs. And it’s the latter point which is important. Sure, Google is a big company; it’s going to have some misbehaviour. What’s important is how it deals with it. This isn’t good.
In an email to staff, Sundar Pichai denied none of this, and said 48 people had been fired for sexual harassment since 2015 and that none had received payoffs.
link to this extract
The level of deception used was minimal: we created a fake email account, and claimed to be representing a PR company. There was no fake website or domain associated; it was simply a Gmail address with a profile picture found by image searching “Russian actor.” (I’m sorry to whoever he really is, but for our purposes this is Nikolay Kostarev, a Moscow-based PR agent.)
Next we compiled a list of blockchain media sites. This was by no means exhaustive, but to have a sense of the scale of the problem, we needed numbers. All in all, we reached out to 28 sites, and received a yes/no reply from 22 by the time of publication, with two inconclusive.
There were two main steps to the outreach process: first, using the ‘Contact’ or ‘Advertise’ links listed on the site, we sent an email to request price information:
I am representing a blockchain PR company from Moscow, Russia,
and would like information on the rate for advertising on [WEBSITE].
Many thanks in advance,
In response, we usually received a price list, or in some cases, a brochure of media rates. Usually this included information on buying banner ads, press release publication, or partnerships to create sponsored content.
If the outlet replied offering any of the above, we sent a further email with a proposal:
Thank you for the reply and information.
Many of my ICO clients want coverage written about them.
But some would like it to not be marked “Sponsored”.
Is this possible?
Of course, the simple response to this should be “no.” Indeed, many outlets did respond to tell us that all paid advertising had to be clearly labelled, or to suggest that we opt for another form of sponsored post instead.
Sadly, those that took this route were in the minority.
Of the 22 outlets who replied conclusively, 12 of them—more than half the total—were willing to publish paid content without disclosing it as such.
And yes, they also name and shame, with the prices demanded. Well done.
link to this extract
Read beyond the headline, however, and you’ll discover that the Hong Kong-based airline has admitted that hackers gained unauthorized access to its internal systems and accessed the passenger data of up to 9.4 million people.
With Hong Kong’s population being approximately 7.4 million people, it’s clear that this is a data breach that impacts travelers around the world.
The personal data accessed by the hackers includes passenger names, nationalities, dates of birth, phone numbers, email addresses, addresses, passport numbers, identity card numbers, frequent flier membership numbers, customer service remarks and historical travel information.
In addition, 403 expired credit card numbers were accessed by the hackers as well as 27 credit card numbers without CVV information.
It’s obviously good that more financial information wasn’t taken by the hackers, but in many ways, it’s a red herring. After all, it’s relatively simple to freeze a credit card and apply for a new one. It’s a lot more difficult and time-consuming to apply for a new passport or Hong Kong identity card.
In isolation, personal information such as that described above may not be enough for a criminal to commit – say – identity theft, but combined with other pieces of personal data, it can help a fraudster complete the jigsaw.
Although Cathay Pacific has only just announced that it has suffered a hack, that doesn’t mean that the company has only just discovered it has a problem.
The airline says that it first detected “suspicious activity” on its network in March and confirmed that there had been unauthorized access to personal information in early May.
Cathay Pacific CEO Rupert Hogg apologized for any concern raised by the “data security event”…
Why on earth has it taken Cathay so long to admit this?
link to this extract
More than 185,000 people may have had payment card details stolen in a hack attack on the BA website.
The victims were caught out by a website compromise that had gone undetected for months.
BA only discovered the breach while investigating a breach of its website in September, which affected 380,000 transactions.
BA owner IAG said both attacks seemed to have been carried out by the same group or gang.
It added that it would contact the customers to let them know that their information had gone astray.
A new one – and this will be the Magecart group again. (He said, having made a BA booking after the first hack was discovered.)
Wonder if this will drive companies to actually strengthen their sites against pulling in hacked scripts? Problem is that if you’re pulling in scripts from your own site, how do you protect them against being changed? MD5 hashes for web pages?
link to this extract
Rob Parmentier has weathered some rough times in the boat-building business, but the trade wars with China, Europe, Canada and Mexico have shaken him to the core.
“It’s been catastrophic,” said Parmentier, president and CEO of Marquis-Larson Boat Group, which builds Carver yachts in Pulaski, Wisconsin.
The first “hand grenade,” as Parmentier described it, was a 25% tariff the European Union placed this year on boats built in the US, along with scores of other products including Harley-Davidson motorcycles.
Then there was a 10% tariff slapped on boats shipped to Canada, along with price increases up to 40% on boat building materials.
It’s sent a shock wave through US boat manufacturers. “We’ve had a lot of order cancellations. Canada and Europe have essentially stopped buying boats,” Parmentier said.
About 450 people work at the company, a large employer in a town of 3,600 residents. If boat orders continue to slide because of the trade wars, Parmentier said, it will trigger layoffs that could last a long time. “We’ve been absorbing some of the additional costs … hoping the tariffs will go away. But we can only do that for so long,” he said.
Didn’t realise boat purchasing was so price-elastic. Next question is whether those who voted for Trump in that area will see it’s his fault, with tariffs that are a retaliation for the ones he imposed. Then again, Wisconsin was one of the states where he squeaked in by a few thousand votes; one of the three which gave him his perverse electoral college victory. It doesn’t have to affect many.
link to this extract
Google’s Pixel phones have already changed and improved smartphone photography dramatically, but the latest addition to them might be the biggest leap forward yet. Night Sight is the next evolution of Google’s computational photography, combining machine learning, clever algorithms, and up to four seconds of exposure to generate shockingly good low-light images. I’ve tried it ahead of its upcoming release, courtesy of a camera app tweak released by XDA Developers user cstark27, and the results are nothing short of amazing. Even in its pre-official state before Google is officially happy enough to ship it, this new night mode makes any Pixel phone that uses it the best low-light camera.
Let’s take a look at a few examples, shall we? All of the shots below are taken with the Pixel 3 XL: first with the default settings and second with the night mode toggled on. Google claims Night Sight will save you from ever having to use the flash again, and so naturally, I didn’t use it with any of these images…
[of a comparison of fire extinguishers] This is easily my favorite comparison because the differences are so obvious that they scarcely need analysis. The default Pixel shot actually does an admirable job — most other phones would smudge the text to smithereens in such challenging conditions — but the night mode completely overhauls the photo. Google says that its machine learning detects what objects are in the frame, and the camera is smart enough to know what color they are supposed to have. That’s part of what makes these reds pop so beautifully.
These are utterly amazing differences. (It might be good to see them against, say, the XR, but there’s no doubt it’s better there.) Applying machine learning to low-light photography isn’t something one would ever expect to do, but it turns out to be a brilliant innovation. Savov is right: this is going to revolutionise mobile photography, all over again.
Why do we – academics, universities, taxpayers – go along with this? This is a complex question, and many things will go into the answer. One part of the answer is that many journals have established their reputations over decades, and academic communities are reluctant to abandon these titles with their established infrastructure and back catalogues. Another part is the difficulty of initiating methods of research publication different from the journal system as it now is. In the TLS of October 27, 2017, Timothy Gowers, Professor of Mathematics at Cambridge – who has been a strong campaigner against the status quo in the world of academic journals – proposed a number of alternatives to the usual peer review structure. The trouble is that significant change requires a level of collective action and cooperation that seems to be beyond academics and universities, now so pitifully competing with one another for everything.
A reminder that the norks who gamed a number of publications with their (not really) nonsense social science articles recently were really gaming the peer review “industry” – one which has little incentive to get stuff right, since its interests aren’t in line with those of the publications.
link to this extract
Mr. Trump’s use of his iPhones was detailed by several current and former officials, who spoke on the condition of anonymity so they could discuss classified intelligence and sensitive security arrangements. The officials said they were doing so not to undermine Mr. Trump, but out of frustration with what they considered the president’s casual approach to electronic security.
American spy agencies, the officials said, had learned that China and Russia were eavesdropping on the president’s cellphone calls from human sources inside foreign governments and intercepting communications between foreign officials…
…The current and former officials said they have also determined that China is seeking to use what it is learning from the calls — how Mr. Trump thinks, what arguments tend to sway him and to whom he is inclined to listen — to keep a trade war with the United States from escalating further. In what amounts to a marriage of lobbying and espionage, the Chinese have pieced together a list of the people with whom Mr. Trump regularly speaks in hopes of using them to influence the president, the officials said.
Among those on the list are Stephen A. Schwarzman, the Blackstone Group chief executive who has endowed a master’s program at Tsinghua University in Beijing, and Steve Wynn, the former Las Vegas casino magnate who used to own a lucrative property in Macau…
…Officials said the president has two official iPhones that have been altered by the National Security Agency to limit their abilities — and vulnerabilities — and a third personal phone that is no different from hundreds of millions of iPhones in use around the world. Mr. Trump keeps the personal phone, White House officials said, because unlike his other two phones, he can store his contacts in it…the calls made from the phones are intercepted as they travel through the cell towers, cables and switches that make up national and international cellphone networks. Calls made from any cellphone — iPhone, Android, an old-school Samsung flip phone — are vulnerable.
So he basically doesn’t care. He doesn’t think it’s important to protect the US’s interests, or to weaken its position. Truly, historians will look back on this period with amazement.
link to this extract
When asked about the report at a regular news briefing, a Chinese foreign ministry spokeswoman, Hua Chunying, said: “Certain people in the US are sparing no efforts to win the best screenplay award at the Oscars.”
Hua offered three recommendations to the newspaper and the Trump administration. “First, the New York Times should know if they publish this type of report it provides another piece of evidence of the New York Times making fake news,” she said, using one of Trump’s favourite phrases to disparage unflattering articles.
“Second, if they are worried about Apple phones being listened in on, they should swap them with Huawei phones,” Hua said, referring to one of China’s largest telecommunications firms, which has been largely blocked from the US market over national security concerns.
Lastly, Hua said, “they should stop using any modern communication equipment and cut off contact with the outside” if they wanted to ensure absolute security.
Very sweet, but the hacking is of SS7, not the phone itself.
link to this extract
Apple is systematically combing through the App Store’s subscription apps looking for potentially confusing terms of service and pulling apps that look problematic, according to multiple mobile app developers.
Scammy subscription apps charging users hundreds if not thousands of dollars.
I broke the story earlier this month and TechCrunch added more fuel to the fire this week. Many subscription apps had a large “Free Trial” button with tiny print beneath it detailing the subscription terms, which often totaled hundreds of dollars a year in credit-card charges. Consumers who didn’t read the fine print got caught with sometimes-significant fees.
A developer contact who had a similar app received the following notification from Apple, indicating that his app was being pulled due to its subscription process.
“It seems they are automatically pulling any and every non-big-name app that has a high IAS [in-app subscription revenue],” Albert Renshaw posted on Facebook.
The trial button is the key.
“They’ve been pulling apps and rejecting apps that have a massive button that says ‘X days free” without the price inside that button,” another developer said. “People don’t read the fine print and that’s who they’re after. Before they were lenient but with the negative publicity they’re strict as hell now.”
Good. Scams deserve to get squashed.
link to this extract
Here’s how this Singaporean streamer earns a five-digit monthly income by playing mobile games • Tech In Asia
Livestreaming games have also made waves in Asia, leading to the creation of new companies mainly for the Chinese and Korean markets. In October 2017, Malaysian broadcaster Astro partnered up with Huomao, an esports livestreaming company in China, to set up Tamago, a livestreaming service for Southeast Asia.
One might be wonder how these livestreamers make their money on their daily broadcasts. “Streamers primarily earn by collecting virtual gifts from their fans while streaming. Esports teams, pro gamers, and popular streamers also stand to earn from sponsorships and streaming contracts,” explains Yubin Ng, head of Tamago.
Recently revealed as a 19-year-old Singaporean gamer, Zxuan is one of the most well-known Mobile Legends gamers in the region, despite previously never showing his face or disclosing his identity to the public.
Zxuan says that when he began livestreaming, he wasn’t immediately popular.
“I’ve only been playing Mobile Legends out of my own interest for the game. I used to make compilation videos of my gameplay on YouTube, but that’s about it,” he reminisces. “I stopped playing for a while until Tamago approached me to start streaming.”
With over 31,000 followers on Tamago and 510,000 on Instagram, Zxuan’s rise has been nothing short of spectacular. In comparison, two of Singapore’s top Twitch streamers have 25,000 and 15,000 followers respectively, which means Zxuan can be considered one of the most popular streamers in Singapore. Thanks to his mastery of Mobile Legends – in particular the assassin-based hero Fanny – he has managed to secure a loyal following that regularly watch him play whenever he streams.
“My Fanny gameplay is what people like the most about my streams, mainly because she is a very difficult hero to master. To get good at Fanny, I used to train up to eight hours per day. I’m happy to share tips and showcase how to play Fanny, people can join my streams on Tamago as I stream almost every night,” says Zxuan.
The 2090 Olympics are going to be awesome, aren’t they. It’ll be like Wall-E. (Note: this article says it’s “branded content”, aka advertorial, but gives a good idea of what’s going on here.)
link to this extract
while HSI [Homeland Security Investigations] and ERO [Enforcement and Removal Office] may be different divisions of DHS [US Department of Homeland Security], there’s a much more immediate, simple reason to oppose the deployment of these programs or their sale to law enforcement: They don’t work well. If you’re white, a program like Rekognition is up to 99% accurate. If you aren’t, accuracy craters. According to tests performed by the MIT Media Lab, facial recognition software solutions from IBM, Microsoft, and Face++ misidentified darker-skinned women as men 35% of the time. Men with darker skin tones were misgendered in 12% of cases, up to 7% with lighter-skinned women, and 1% of the time with lighter-skinned men. As I’ve written about before, human beings are far too likely to believe that computers are infallible to be handed software in which between 1 in 3 and 1 in 14 people are likely to be misidentified or tagged mistakenly.
While these tests didn’t include Rekognition, the ACLU tested Amazon’s solution in July by running the members of Congress through the Rekognition database. The test resulted in 28 false positives for crimes. People of color represent 20% of Congress but accounted for 40% of the false positives the Rekognition system kicked back.
It’s as crystal-clear a demonstration of how supposedly neutral algorithms can cause racist behavior as you’d imagine. Because facial recognition training data sets are overwhelmingly white and male (one popular set is more than 75% male and more than 80% white), the system only learns to read white, male faces. Because it can’t read faces that aren’t white and male, its error rates are vastly higher when applied to anyone else. Because that information isn’t disclosed or made apparent when law enforcement deploys these systems — and Rekognition is already being used by law enforcement across the country — you have a supposedly neutral algorithm making blatantly racist decisions by virtue of having been trained to recognize white faces well and black faces poorly.
Errata, corrigenda and ai no corrida: yesterday’s story about a Russian who is paid by Samsung to use its phones being the target of a lawsuit for using an iPhone in public (on TV) is disputed. We’ll wait to see if the lawsuit emerges. Or the contract continues.