for years, CAPTCHAs have fed Google’s AI systems with valuable data – but now Cloudflare has a quicker alternative. CC-licensed photo by Becky Stern 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 10 links for you. No squiggly text. I’m @charlesarthur on Twitter. Observations and links welcome.
Cloudflare’s CAPTCHA proposal would end AI’s source of free labor • Quartz
Nicolás Rivero:
»
Are you a human? If so, chances are you’ve filled out countless CAPTCHAs, the all-too-frequent tests internet users have to take to prove their humanity by identifying garbled text, fuzzy numbers, or images of traffic lights.
The tests serve the crucial function of differentiating genuine human web users from malign bots attempting to hack or spam a website. But they’re also annoying and time-consuming. Web infrastructure company Cloudflare estimates humanity collectively spends 500 years of labor each day on CAPTCHAs. In a May 13 blog post, the company declared its intention to “get rid of CAPTCHAs completely” through alternatives that wouldn’t require people to complete arbitrary tasks.
These alternatives aren’t entirely new. Identity verification firm Yubico has been selling flash drives that web users can use to prove their humanity since 2008, and Google launched a “No CAPTCHA” technique in 2014 that can confirm some web users’ humanity just by monitoring how they interact with webpages. Cloudflare stopped using Google’s CAPTCHA service last year after the search giant began charging for it, which eventually led to Cloudflare’s recent commitment to reinventing the CAPTCHA.
Alternative solutions, however, will prove very disruptive for the other, less public purpose of CAPTCHA tests: they’re a massive source of free labour for AI developers. Killing the CAPTCHA would derail the gravy train that has provided cheap advances in the field of machine vision for the past decade.
«
Though as the story makes more clear, CAPTCHAs have been a massive source of free labour for *Google’s* AI developers. Not anyone else. Facebook doesn’t use them. Cloudflare’s solution, for now, is a hardware key. (Its reasons for hating CAPTCHAs – particularly their implicit cultural imperialism – are worth reading.)
unique link to this extract
Israel-Gaza: Why is the region blurry on Google Maps? • BBC News
Christopher Giles and Jack Goodman:
»
Why is Gaza, one of the most densely populated places in the world, blurry on Google Maps?
…on Google Earth, the most widely used image platform, the most recent imagery for Gaza is of low resolution and therefore blurry. “The most recent Google Earth image is from 2016 and looks like trash. I zoomed in on some random rural area of Syria and it has had 20+ images taken since that time, in very high resolution,” tweeted Aric Toler, a journalist for Bellingcat.
…Until last year, the US government restricted the quality of satellite images that American companies were permitted to provide on a commercial basis. The Kyl-Bingaman Amendment (KBA) had been introduced in 1997 to address Israeli security concerns.
Although the ruling only referred to Israel, it was also applied the restriction to images of the Palestinian territories. The KBA limited image quality so that an object the size of a car was just about visible as a highly blurred image, but anything smaller would be very difficult to identify.
“We [Israel] would always prefer to be photographed at the lowest resolution possible”, said Amnon Harari, head of space programmes at Israel’s Defence Ministry last year, reported by Reuters. “It’s always preferable to be seen blurred, rather than precisely.”
«
They’re being updated, but it could take a while.
unique link to this extract
Censorship, surveillance and profits: a hard bargain for Apple in China • The New York Times
Jack Nicas, Raymond Zhong and Daisuke Wakabayashi:
»
The Chinese government regularly demands data from Chinese companies, often for law-enforcement investigations. Chinese law requires the companies to comply.
US law has long prohibited American companies from turning over data to Chinese law enforcement. But Apple and the Chinese government have made an unusual arrangement to get around American laws.
In China, Apple has ceded legal ownership of its customers’ data to Guizhou-Cloud Big Data, or GCBD, a company owned by the government of Guizhou Province, whose capital is Guiyang. Apple recently required its Chinese customers to accept new iCloud terms and conditions that list GCBD as the service provider and Apple as “an additional party.” Apple told customers the change was to “improve iCloud services in China mainland and comply with Chinese regulations.”
The terms and conditions included a new provision that does not appear in other countries: “Apple and GCBD will have access to all data that you store on this service” and can share that data “between each other under applicable law.”
Under the new setup, Chinese authorities ask GCBD — not Apple — for Apple customers’ data, Apple said. Apple believes that gives it a legal shield from American law, according to a person who helped create the arrangement. GCBD declined to answer questions about its Apple partnership.
«
OK, China is an authoritarian government. Its citizens live under an authoritarian regime. Question is, does Apple’s presence there help the government? If not (and I’d say it doesn’t), does using Apple’s products help dissidents evade the regime? (Probably not – they don’t need to crack the phone if they can crack your head.) Google withdrew because the state was hacking its product to target opponents. There’s no evidence that Apple’s been subverted in the same way. This long piece essentially reiterates what we’ve known for a long time: there’s no perfect way to interact with China. It always involves moral compromise. Even *not* interacting means you’re not helping.
unique link to this extract
AirPods Max and AirPods Pro don’t support Apple Music Lossless, Apple confirms • T3
Matthew Bolton:
»
Apple has announced that it’s adding ‘Lossless’ and ‘Hi-Resolution Lossless’ streaming options to Apple Music in June 2021 for no extra charge, as well as offering Dolby Atmos ‘Spatial Audio’ 3D music, too.
In Apple’s new terminology, ‘Lossless’ is CD quality, from 16-bit 44.1kHz playback up to 24-bit 48kHz, while ‘Hi-Res Lossless’ delivers up to 24-bit 192kHz. Don’t worry if you don’t know what that means – it means music comes in larger files with much less compression, meaning more realistic results, provided you’ve got good enough equipment to actually hear the difference.
Apple has confirmed to T3 that this equipment, sadly, does not include AirPods Pro or AirPods Max. Both of Apple’s elite headphone models only use the Bluetooth AAC codec when connected to an iPhone, which means they can’t receive the full quality of the Apple Music ‘Lossless’ files, which will be encoded as ALAC (Apple Lossless Audio Codec) files.
What both of these devices will be able to receive is the new Dolby Atmos ‘Spatial Audio’ versions of songs, which will add more of a surrounding 3D effect in tracks. These aren’t the only headphones that support this feature – anything powered by Apple’s H1 or W1 wireless chips will, and that includes (deep breath): AirPods, AirPods Pro, AirPods Max, BeatsX, Beats Solo3 Wireless, Beats Studio3, Powerbeats3 Wireless, Beats Flex, Powerbeats Pro, and Beats Solo Pro.
«
The AirPods Max won’t even do lossless over the Lightning charging cable (which is an option for playing audio). A firmware update might fix that, but you’d expect that “group working on audio hardware” and “group working on audio software” might talk to each other about future plans? Taken with the hiring and then firing of Antonio García Martinez, it feels like the gaps in Apple’s internal culture are showing. (Side note: Amazon is also making lossless available on Amazon Music, also for free.)
unique link to this extract
How Substack soap operas change the media business • The Atlantic
Helen Lewis:
»
Normal people—with regular lives and real jobs—have soap operas and reality shows. People who are Extremely Online have Substack.
Over the past few months, the PR travails of the newsletter start-up have become a reliable source of media gossip. Jude Doyle is leaving! Grace Lavery has joined! Oh man, Matt Yglesias shouldn’t have taken that advance; he’d have made far more money purely from subscriptions!
Perhaps those names don’t mean anything to you. Why should they? Doyle has 43,000 Twitter followers, a fan base 20 times smaller than that of the Sarcastic Mars Rover parody account. Lavery is an English professor, an expert on Japanese Victoriana, and one-third of a Brooklyn throuple that also includes Daniel Lavery, who has a Substack named after William Shatner. (Together, the Laverys have received $555,000 in advances from the platform.) Yglesias was an old-school blogger, then co-founded Vox, and has now returned to his independent roots.
But for a certain subset of the American elite—a group of people who are concentrated in journalism, academia, and related fields; who are likely to be active on Twitter; and who have strong opinions on the 1619 Project and the ACLU’s Chase Strangio—following the lives of these people is what they do instead of watching General Hospital or The Bachelor. Many of the authors now showing up on Substack are known for fighting with journalists at other outlets, and one another. By supporting their newsletters, readers get endless feuds, dramatic exits, high-profile guest stars, ambitious crossover events, and compelling villains. Yes, Substack is selling soap operas to people who think they’re above soap operas.
«
Lewis does have her own Substack, but it’s free (and funny), and she doesn’t mess around with internet beefs. Though of course she does comment, when linking to this article in her most recent one, that “You will be pleased to know that everyone I lightly ribbed in this piece responded with the self-deprecating humour and good nature for which they are famed.” 😂
unique link to this extract
Google I/O 2021 preview: Google resurrects Wear OS and Android tablets? • Ars Technica
Ron Amadeo, on Google’s event which starts today, Tuesday:
»
The biggest sign that Google is bringing Wear OS back to the land of the living is a widely reported rumor that the Samsung Galaxy Watch 4 will run Wear OS instead of Tizen. Plugged-in Samsung leaker Ice Universe called the transition back in February, kicking off this batch of rumors. The latest report from the Korean site MT says Samsung wants to switch to Wear OS due to difficulty in getting developers to create Tizen apps. XDA Developers found references to a “Merlot Wear OS” device in a Samsung Wi-Fi driver, indicating the company is at least experimenting with Wear OS internally. Samsung has been kicking around the idea of returning to Wear OS for some time, though, but the company has yet to pull the trigger.
Samsung adopting Wear would solve a lot of problems. Samsung makes its own smartwatch chips, so the platform could finally stop relying on Qualcomm for smartwatch SoCs. Qualcomm has smothered Wear OS with a lack of significant chip upgrades, which greatly contributed to the current situation. Samsung is also a top-tier hardware manufacturer, so it can push the smartwatch form factor forward with whatever parts it wants. The fashion brands that occupy the Wear OS market right now (like Fossil) can really only source existing parts.
There’s also that $2.1bn acquisition of Fitbit, which Google closed in January. There have been some product launches since then, but we’ve yet to see what the Googlification of Fitbit looks like.
…The next long-dead Android form factor that has been suddenly active lately is the Android tablet. Google’s tablet interfaces for Android probably peaked around Android 3 or 4.0 when tablets first came out. Since then, the company has consistently removed and scaled back tablet interfaces. Google’s hardware division also hasn’t made an Android tablet for years, with the last release being the Pixel C in 2015.
Google’s lack of interest in tablets seems to be changing, though. Google surprised us all earlier this month with the announcement of “Entertainment Space” for tablets, a media aggregation UI that will appear as a home screen panel, replacing the usual Google Discover feed on the left side.
«
Pretty sure that Android tablets are not going to become A Thing, at least in terms of third-party apps. Same problem as Samsung/Tizen: no traction.
unique link to this extract
Ransomware’s dangerous new trick is double-encrypting your data • WIRED
Lily Hay Newman:
»
Double-encryption attacks have happened before, usually stemming from two separate ransomware gangs compromising the same victim at the same time. But antivirus company Emsisoft says it is aware of dozens of incidents in which the same actor or group intentionally layers two types of ransomware on top of each other.
“The groups are constantly trying to work out which strategies are best, which net them the most money for the least amount of effort,” says Emsisoft threat analyst Brett Callow. “So in this approach you have a single actor deploying two types of ransomware. The victim decrypts their data and discovers it’s not actually decrypted at all.”
Some victims get two ransom notes at once, Callow says, meaning that the hackers want their victims to know about the double-encryption attack. In other cases, though, victims only see one ransom note and only find out about the second layer of encryption after they’ve paid to eliminate the first.
“Even in a standard single-encryption ransomware case, recovery is often an absolute nightmare,” Callow says. “But we are seeing this double-encryption tactic often enough that we feel it’s something organizations should be aware of when considering their response.”
Emsisoft has identified two distinct tactics. In the first, hackers encrypt data with ransomware A and then re-encrypt that data with ransomware B. The other path involves what Emsisoft calls a “side-by-side encryption” attack, in which attacks encrypt some of an organization’s systems with ransomware A and others with ransomware B. In that case, data is only encrypted once, but a victim would need both decryption keys to unlock everything.
«
This seems like a strategic mistake by the ransomware types: if there’s a possibility that you might get double-crossed even if you pay, then people will be less likely to pay.
unique link to this extract
The app that lets you pay to control another person’s life • BBC News
Will Smale:
»
How would you feel about being able to pay to control multiple aspects of another person’s life? A new app is offering you the chance to do just that.
When writer Brandon Wong recently couldn’t decide what takeaway to order one evening, he asked his followers on social media app NewNew to choose for him.
Those that wanted to get involved in the 24-year-old’s dinner dilemma paid $5 (£3.50) to vote in a poll, and the majority verdict was that he should go for Korean food, so that was what he bought.
“I couldn’t decide between Chinese or Korean, so it was very helpful,” says Mr Wong, who lives in Edmonton, Canada. “I have also used NewNew polls to decide what clothes I should wear that day, and lots of other personal stuff.
“I joined back in March, and I post [polls] three or four times a week. I’ve now had more than 1,700 total votes.”
NewNew is the brainchild of Los Angeles-based entrepreneur Courtne Smith. The app, which is still in its “beta” or pre-full release stage, describes itself as “a human stock market where you buy shares in the lives of real people, in order to control their decisions and watch the outcome”.
«
This feels more like the setup for a Philip K Dick short story, in which the writer then discovers that he is actually an android who really is being controlled by humans. (That’s essentially his story The Electric Ant.) But no, it’s Silicon Valley.
unique link to this extract
Bill Gates left Microsoft board amid probe into prior relationship with staffer • WSJ
Emily Glazer, Justin Baer, Khadeeja Safdar and Aaron Tilley:
»
Microsoft Corp. board members decided that Bill Gates needed to step down from its board in 2020 as they pursued an investigation into the billionaire’s prior romantic relationship with a female Microsoft employee that was deemed inappropriate, people familiar with the matter said.
Members of the board tasked with the matter hired a law firm to conduct an investigation in late 2019 after a Microsoft engineer alleged in a letter that she had a sexual relationship over years with Mr. Gates, the people said.
During the probe, some board members decided it was no longer suitable for Mr. Gates to sit as a director at the software company he started and led for decades, the people said. Mr. Gates resigned before the board’s investigation was completed and before the full board could make a formal decision on the matter, another person familiar with the matter said.
…A spokeswoman for Mr. Gates said, “There was an affair almost 20 years ago which ended amicably.” She said his “decision to transition off the board was in no way related to this matter. In fact, he had expressed an interest in spending more time on his philanthropy starting several years earlier.”
Mr. Gates resigned from the Microsoft board on March 13, 2020, three months after he had been re-elected to his seat.
«
That he wasn’t a delightful person to work for – yelling and screaming at subordinates (ie pretty much everyone) – was already known; but there’s a darker, or perhaps seamier, side coming out now. Melinda seems to have been very unhappy about Bill seeing Jeffrey Epstein. There’s more to come on that.
unique link to this extract
In Mahle’s contact-free electric motor, power reaches the rotor wirelessly • IEEE Spectrum
Philip Ross:
»
Designs that put copper windings in the rotor have to transmit electricity to a moving target, and the point of contact—the slip ring—is subject to wear and tear.
Mahle, a German auto parts company, unveiled a motor that’s free of both rare earths and of physical contact. Power is beamed into the rotor wirelessly, through induction, by a coil carrying alternating current. This induces a current in the receiving electrode, inside the rotor, which energizes the copper windings there to produce an electromagnetic field.
That means there’s practically nothing that can wear out. “There are no contacts to transmit electricity, no abrasion, no dust formation, no mechanical wear,” Martin Berger, Mahle’s head of research, said Wednesday, in an online press conference. “Also I have to say, if one must service a non-magnetized rotor, it’s not difficult to exchange the rotor.”
It may seem strange to try to minimize wear and tear in electric motors, seeing as they are already famed for their simplicity and durability. Unlike internal-combustion engines, electric motors have practically no moving parts, and they are fairly easy to take apart and put back together. Perhaps Mahle’s engineers got the idea from their longstanding work in wireless charging technology. Maybe the contact-free rotor design provides advantages beyond mere durability.
Berger says the new motor combines the best points of several motor designs, for instance by offering good efficiency at both low and high torque. Overall, the company asserts, the motor achieves at least 95% efficiency in typical EV use and tops 96% efficiency at many operating points. A release from Mahle says that no EV except for Formula E racing cars has done better.
«
The absence of rare earths means no (or less) reliance on China. Commercial production perhaps two or three years away.
unique link to this extract
Errata, corrigenda and ai no corrida: none notified
Hey, you made it this far – why not round it all off by preordering Social Warming, my forthcoming book?
The Overspill: when there's more that I want to say 