Start Up No.1328: Amazon suspends facial recognition system, Zoom blocks user over Tiananmen party, life in the US police, and more

Everybody is doomscrolling – flicking endlessly through screens, glued to what’s unfolding CC-licensed photo by verchmarco on Flickr.

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 12 links for you. Read them all! I’m @charlesarthur on Twitter. Observations and links welcome.

Amazon suspends police use of its facial-recognition technology • WSJ

Asa Fitch:


Amazon Inc said it is halting law-enforcement use of its facial-recognition technology for a year following budding congressional efforts to regulate such tools amid widespread criticism about racial and gender bias.

“We hope this one-year moratorium might give Congress enough time to implement appropriate rules, and we stand ready to help if requested,” Amazon said in a blog post Wednesday. The retailing giant said it has been advocating for strong government regulation of the use of facial-recognition technology, and Congress appeared ready to take on that challenge.

A police reform bill House Democratic lawmakers introduced Monday would prohibit federal law enforcement’s use of real-time facial recognition.

Amazon has sold its Rekognition face-recognition software widely, including to police departments and other U.S. enforcement agencies. The company said it would continue to allow the use of its tools by organizations that deploy facial recognition to combat human trafficking and find missing children.


So the police won’t have it. But read on…
unique link to this extract

This simple facial recognition search engine can track you down across the internet • OneZero

Dave Gershgorn:


Ever wondered where you appear on the internet? Now, a facial recognition website claims you can upload a picture of anyone and the site will find that same person’s images all around the internet.

PimEyes, a Polish facial recognition website, is a free tool that allows anyone to upload a photo of a person’s face and find more images of that person from publicly accessible websites like Tumblr, YouTube, WordPress blogs, and news outlets.

In essence, it’s not so different from the service provided by Clearview AI, which is currently being used by police and law enforcement agencies around the world. PimEyes’ facial recognition engine doesn’t seem as powerful as Clearview AI’s app is supposed to be. And unlike Clearview AI, it does not scrape most social media sites.

PimEyes markets its service as a tool to protect privacy and the misuse of images. But there’s no guarantee that someone will upload their own face, making it equally powerful for anyone trying to stalk someone else. The company did not respond to a request for comment.

PimEyes monetizes facial recognition by charging for a premium tier, which allows users to see which websites are hosting images of their faces and gives them the ability to set alerts for when new images are uploaded. The PimEyes premium tiers also allow up to 25 saved alerts, meaning one person could be alerted to newly uploaded images of up to 25 people across the internet. PimEyes has also opened up its service for developers to search its database, with pricing for up to 100 million searches per month.

Facial recognition search sites are rare but not new. In 2016, Russian tech company NtechLab launched FindFace, which offered similar search functionality, until shutting it down in a pivot to state surveillance. Its founders described it as a way to find women a person wanted to date.


And now it’s a way to find a person who’s wanted from a particular date! Brm-tish! Oh, suit yourselves. The lack of social media input on this makes it not that much different from many other image searches. And it will definitely get used for stalking.
unique link to this extract

Why Facebook staffers won’t quit over Trump’s posts • The Atlantic

Ian Bogost:


It’s easier for tech workers to talk about taking a stand than to do so. For one, big technology companies such as Facebook and Google are viciously competitive about acquiring talent. They hire or poach the best people, sometimes just to prevent a competitor from having access to them instead. Some workers don’t want to rock the boat for fear they might get blacklisted, McCarthy said. And ironically, the brokenness at companies such as Facebook and Uber can also make their jobs enticing. Disruption is appealing, and the promise to move fast and break things (even priceless and irrecoverable ones, such as democracy) can be a recruiting tool.

Others already in a company’s employ may see an opportunity to fix some of its ills. One product manager at a large tech firm, who also advises many early-career professionals, spoke with me on the condition of anonymity because she fears reprisal from within the industry. She told me about her “activist” friends who refuse to leave jobs at Facebook, even if they disagree with the company’s practices. “They came to change the world,” she said, “and stayed to work within the system on issues they cared about.” The same drive that makes these workers care about the consequences of Facebook’s impact on democracy also makes them want to stick it out in an effort to improve the service.

Even so, Facebook seems to have crossed the line of tolerable abhorrence for some tech workers. Inside the business, nextplayism may offer the best, and maybe the only, way for them to show their distaste. “The vast majority of people I know at the director-and-up level, when they are leaving a company and looking for a new gig, they’re Never Facebookers,” McCarthy, who is also an occasional collaborator of mine, said, referring to senior-level roles. “They’re offended if you even offer to do introductions to someone at Facebook.”

But that is a privileged attitude. Much of the magical operation of online services is driven by rote laborers, such as moderators, AI-training wranglers, and gig workers. They aren’t counted as members of the industry, except perhaps as its casualties.


unique link to this extract

Facebook helped the FBI hack a child predator • VICE

Lorenzo Franceschi-Bicchierai:


For years, a California man systematically harassed and terrorized young girls using chat apps, email, and Facebook. He extorted them for their nude pictures and videos, and threatened to kill and rape them. He also sent graphic and specific threats to carry out mass shootings and bombings at the girls’ schools if they didn’t send him sexually explicit photos and videos.

Buster Hernandez, who was known as “Brian Kil” online, was such a persistent threat and was so adept at hiding his real identity that Facebook took the unprecedented step of helping the FBI hack him to gather evidence that led to his arrest and conviction, Motherboard has learned. Facebook worked with a third-party company to develop the exploit and did not directly hand the exploit to the FBI; it is unclear whether the FBI even knew that Facebook was involved in developing the exploit. According to sources within the company, this is the first and only time Facebook has ever helped law enforcement hack a target.

This previously unreported case of collaboration between a Silicon Valley tech giant and the FBI highlights the technical capabilities of Facebook, a third-party hacking firm it worked with, and law enforcement, and raises difficult ethical questions about when—if ever—it is appropriate for private companies to assist in the hacking of their users. The FBI and Facebook used a so-called zero-day exploit in the privacy-focused operating system Tails, which automatically routes all of a user’s internet traffic through the Tor anonymity network, to unmask Hernandez’s real IP address, which ultimately led to his arrest.


Facebook felt it had no choice: that this wasn’t an encryption backdoor, that he represented a unique threat, that there weren’t wider privacy risks. Still controversial inside the company, though. This happened two years ago.
unique link to this extract

Mark Zuckerberg is on the wrong side of history • Fast Company

Maelle Gavet:


last week when I saw Mark Zuckerberg on Fox News defending Facebook’s laissez-faire approach to the content that populates its site, I thought, “Ok, this is it. Surely when they see their boss on the side of bigotry, hate, and racism, they are going to realize that they are on the wrong side of history.”

So, I called my friends at Facebook to ask them how they were feeling, and to see if they needed to vent. But instead of expressing doubt about the company’s position, most of them doubled down, telling me that “Mark is really the only grownup,” that Twitter is acting irresponsibly by “censoring” President Trump, and that free speech is fundamental—too essential to democracies for Facebook to stifle it.

As I think about these phone calls, it is painfully obvious to me that future dinner parties with these pals will get heated. So, I put together my own cheat sheet to keep in my back pocket for heated conversations to come. Here are my friends’ claims (in italics) and my responses:


There are a lot of points, too long to excerpt, on being the arbiter of truth, “letting people decide”, “ending free speech”, Section 230, fact-checking, “all sides”, bias, and “distractions”. It’s thorough. (Gavet is a CEO in the tech industry.)
unique link to this extract

Confessions of a former bastard cop – Officer A. Cab • Medium

Officer A. Cab:


let me tell you about an extremely formative experience: in my police academy class, we had a clique of around six trainees who routinely bullied and harassed other students: intentionally scuffing another trainee’s shoes to get them in trouble during inspection, sexually harassing female trainees, cracking racist jokes, and so on. Every quarter, we were to write anonymous evaluations of our squadmates. I wrote scathing accounts of their behavior, thinking I was helping keep bad apples out of law enforcement and believing I would be protected. Instead, the academy staff read my complaints to them out loud and outed me to them and never punished them, causing me to get harassed for the rest of my academy class. That’s how I learned that even police leadership hates rats. That’s why no one is “changing things from the inside.” They can’t, the structure won’t allow it.

And that’s the point of what I’m telling you. Whether you were my sergeant, legally harassing an old woman, me, legally harassing our residents, my fellow trainees bullying the rest of us, or “the bad apples” illegally harassing “shitbags”, we were all in it together. I knew cops that pulled women over to flirt with them. I knew cops who would pepper spray sleeping bags so that homeless people would have to throw them away. I knew cops that intentionally provoked anger in suspects so they could claim they were assaulted. I was particularly good at winding people up verbally until they lashed out so I could fight them. Nobody spoke out. Nobody stood up. Nobody betrayed the code.

…I want to highlight this: nearly everyone coming into law enforcement is bombarded with dash cam footage of police officers being ambushed and killed. Over and over and over. Colorless VHS mortality plays, cops screaming for help over their radios, their bodies going limp as a pair of tail lights speed away into a grainy black horizon. In my case, with commentary from an old racist cop who used to brag about assaulting Black Panthers.

To understand why all cops are bastards, you need to understand one of the things almost every training officer told me when it came to using force:


Impossible to be certain if it’s true, of course, but it’s persuasive in its low-key attitude.
unique link to this extract

I was a white woman driving a car. Why did the police keep pulling me over? • FranklyWrite

Cynthia Franks:


I have not told this story before. I worry how it will be received. I don’t know the right language to express it other than my own thoughts and feelings. This post is not for people of color because they already know it. This is for white people living in suburbs and small towns who think this is a big city problem and “It’s not my town.”

Before moving to New York City, I drove every where. I got pulled over 3 times in 15 years; two speeding tickets and an illegal left hand turn.

The first year I was back in Michigan, I got pulled over 5 times. Each time it was for impeding traffic and I did not get a ticket.

I drove a dark grey, 1998 Chevy Venture van that was in storage for several years. It was in good shape.


I’ve changed the headline on this story because it’s so essential that you don’t know the twist. I read much of the story not knowing the writer’s name or colour, and thinking “she” was a black man. Read like that, it became even weirder. But not as weird as the reality.
unique link to this extract

Zoom closes account of U.S.-based Chinese activist after Tiananmen event • Axios

Bethany Allen-Ebrahimian:


Zhou Fengsuo, founder of the U.S. nonprofit Humanitarian China and former student leader of the 1989 Tiananmen protests, organized the May 31 event held through a paid Zoom account associated with Humanitarian China.

About 250 people attended the event. Speakers included mothers of students killed during the 1989 crackdown, organizers of Hong Kong’s Tiananmen candlelight vigil, and others.

On June 7th, the Zoom account displayed a message that it had been shut down, in a screenshot viewed by Axios. Zhou has not been able to access the account since then, and Zoom has not responded to his emails, he told Axios.

A second Zoom account belonging to a pro-democracy activist, Lee Cheuk Yan, a former Hong Kong politician and pro-democracy activist, was also closed in late May. Lee has also received no response from Zoom.


Zoom said it “had to comply with local law” – the people were outside China. But it reopened his account. Doesn’t look good at all.
unique link to this extract

Doomscrolling: why we just can’t look away • WSJ

Nicole Nguyen:


I fixated on the glow in my hand, lighting up an otherwise dark bedroom. In the past few months, after-hours screen time had become a ritual. Last night—and the night before, and the night before that—I stayed up thumbing through tweets, grainy phone-captured videos, posts that gave me hope and posts that made me enraged. I felt like I needed to see it. All of it.

I was “doomscrolling.” Also known as “doomsurfing,” this means spending inordinate amounts of time on devices poring over grim news—and I can’t seem to stop. My timeline used to be a healthy mix of TikTok memes and breaking-news alerts. Now the entire conversation is focused on two topics: the pandemic and the protests.

People are logging on to keep up with it all. This past week, as demonstrations swept the globe, videos from the protests garnered millions of views on social-media platforms. One compilation has been watched more than 50 million times. For the quarter that ended in March, Twitter reported a 24% increase in daily active users over the same period last year. On June 2, Twitter ranked No. 7 in Apple’s App Store—above Facebook, Instagram, Messenger and Snapchat.

On April 24, Merriam-Webster added “doomscrolling” to its “Words We’re Watching” list but the term has circulated since at least 2018. For many, myself included, it has become an irresistible urge, in part because we’re stuck at home, spending too much time on our screens, and in part because that’s precisely where social media’s power over us is amped up.

This has a lot to do with our primal instincts, say experts. Our brains evolved to constantly seek threats—historically, that might mean poisonous berries or a vicious rival tribe, explains Mary McNaughton-Cassill, professor of clinical psychology at the University of Texas at San Antonio. “That’s why we seem predisposed to pay more attention to negative than positive things,” she says. “We’re scanning for danger.”


Well, there’s a lot of doom about, after all.

unique link to this extract

Why Twitter didn’t label Trump’s tweet on Martin Gugino • The New York Times

Kate Conger:


The simple answer: The tweet did not violate the company’s rules, a spokesman said. What Mr. Trump posted about Mr. Gugino, a peace activist who was still in the hospital recovering from a serious head wound, did not cross into narrow areas of content that the company has staked out for closer scrutiny.

Twitter adds fact-checking labels to tweets that contain misinformation about civic integrity or the coronavirus, and tweets that contain “manipulated media,” like photos or videos that have been doctored to mislead viewers. It also places warnings on tweets from world leaders that violate its policy against promoting violence. Similar tweets from regular users are often removed.

No other content — even offensive or inaccurate claims like the ones Mr. Trump posted about Mr. Gugino — gets a label.

The disconnect between putting labels on some of Mr. Trump’s posts and ignoring arguably more offensive content is indicative of how difficult — and confusing — it will be for the company to more closely moderate what the president and other political figures post.

“This case absolutely illustrates the challenges Twitter is facing right now: How can, and how should, a platform moderate a president who regularly pumps polluted information into the ecosystem?” said Whitney Phillips, an assistant professor at Syracuse University. “No decision, whether it’s to respond or not to respond, will be consequence-free.”

Last month, Twitter began adding labels to Mr. Trump’s tweets. The company fact-checked comments he made about elections and placed a warning label over a tweet in which, it said, Mr. Trump glorified violence.
It was the first time that Twitter had taken any action against Mr. Trump, who has long enjoyed free rein on the platform and used it as his preferred method of lobbing insults against rivals and revving up his supporters.

Twitter’s move was met with anger from Mr. Trump and prominent conservatives, who said the company was censoring their voices. Mr. Trump signed an executive order intended to chip away at legal protections for Twitter and other social media companies. That order is already facing a lawsuit challenging its legality.


Alternative: Twitter has to fact-check the internet.
unique link to this extract

Britain goes coal free as renewables edge out fossil fuels • BBC News

Justin Rowlatt:


Britain is about to pass a significant landmark – at midnight on Wednesday it will have gone two full months without burning coal to generate power.

A decade ago about 40% of the country’s electricity came from coal; coronavirus is part of the story, but far from all.

When Britain went into lockdown, electricity demand plummeted; the National Grid responded by taking power plants off the network. The four remaining coal-fired plants were among the first to be shut down.

The last coal generator came off the system at midnight on 9 April. No coal has been burnt for electricity since. The current coal-free period smashes the previous record of 18 days, 6 hours and 10 minutes which was set in June last year.

The figures apply to Britain only, as Northern Ireland is not on the National Grid. But it reveals just how dramatic the transformation of our energy system has been in the last decade. That the country does not need to use the fuel that used to be the backbone of the grid is thanks to a massive investment in renewable energy over the last decade.

Two examples illustrate just how much the UK’s energy networks have changed. A decade ago just 3% of the country’s electricity came from wind and solar, which many people saw as a costly distraction. Now the UK has the biggest offshore wind industry in the world, as well as the largest single wind farm, completed off the coast of Yorkshire last year.


unique link to this extract

fyi: You can bypass YouTube ads by adding a dot after the domain • webdev



On desktop browsers.

For example, // will occasionally show ads // will not show ads

It’s a commonly forgotten edge case, websites forget to normalize the hostname, the content is still served, but there’s no hostname match on the browser so no cookies and broken CORS – and lots of bigger sites use a different domain to serve ads/media with a whitelist that doesn’t contain the extra dot


Just trying to be helpful, that’s all, not at all to defund YouTube.
unique link to this extract

Errata, corrigenda and ai no corrida: none notified

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.