Lunchtime on the Masaai Mara is a lot quieter with solar-powered electric vehicles – a neat adaptation.CC-licensed photo by Ray in Manila 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 11 links for you. Syntactical. I’m @charlesarthur on Twitter. Observations and links welcome.
How Facebook, Twitter and YouTube are dealing with Taliban takeover • CNN
Diksha Madhok, CNN Business:
»
On Tuesday, Facebook (FB) reiterated its ban on accounts praising, supporting, or representing the Taliban from its platforms, including WhatsApp and Instagram, and said that it would remove “accounts maintained by or on behalf of the Taliban.”
“The Taliban is sanctioned as a terrorist organization under US law and we have banned them from our services under our Dangerous Organization policies,” a company spokesperson said.
…other big social media platforms were less clear about their strategy to deal with content supporting the Taliban.
A Twitter spokesperson said that people in Afghanistan are using the platform to seek help, and the company promised to “remain vigilant” in enforcing its policies, including those that ban content that glorifies violence.
One of the Taliban’s spokesmen, Suhail Shaheen, has an active, unverified account on Twitter with 347,000 followers.
YouTube said Tuesday it will “terminate” accounts run by the Afghan Taliban, clarifying its earlier remarks that suggested the group is not banned from the video platform.
Following repeated questions by CNN, YouTube said in a statement that the company “complies with all applicable sanctions and trade compliance laws, including relevant U.S. sanctions. As such, if we find an account believed to be owned and operated by the Afghan Taliban, we terminate it. Further, our policies prohibit content that incites violence.”
YouTube told CNN that because the Afghan Taliban appears on the Treasury Department’s sanctions list, the platform will not allow accounts controlled by the group.
«
Hilariously, Parler – billing itself as “the nation’s premier free-speech social media platform” – demanded that Twitter ban the Taliban on the basis that it’s “uninterested in the free exchange of ideas”. Although there is a lot to ponder in the tweet by CNN correspondent Donie O’Sullivan: “The former President of the United States is banned from Twitter but the Taliban is not.” (To which a number of people replied: “the Taliban haven’t broken Twitter’s terms of service yet.”)
unique link to this extract
Facebook reveals top posts but still won’t share key data about disinformation • Ars Technica
Tim de Chant:
»
Facebook released its first report today detailing which content it says is widely viewed on the site and Instagram. The report comes as research and news stories have highlighted how misleading posts and outright disinformation can draw intense engagement on the company’s platforms.
Much of the scrutiny has focused on far-right accounts, which according to Facebook’s own tool, CrowdTangle, receive the most engagement—likes, shares, and comments. For example, Kevin Roose, a reporter at The New York Times, uses CrowdTangle to tweet out a list of the “10 top-performing link posts by US Facebook pages every day, ranked by total interactions.” What his experiment has revealed is that, day after day, far-right accounts and pages from the likes of Ben Shapiro, Dan Bongino, and Newsmax appeared on the list, sometimes occupying multiple spots. Critics have pointed to the list as evidence that the platform has become a right-wing media machine.
The Twitter account, Roose said, “drove executives crazy” at Facebook. They felt it was making Facebook look like it favored right-wing accounts. All of that brings us to today.
Facebook released the first of what will be a quarterly “Widely Viewed Content Report”. The report will appear alongside its Community Standards Enforcement Report, an existing release which includes data on hate speech and child endangerment. The newest report is Facebook’s attempt to provide more “transparency and context,” said Anna Stepanov, the company’s director of product management.
«
The top 20 viewed content links for the quarter are not at all right wing, and make an odd comparator with the top 10 engagement posts. One feels that something’s just not right about this. Interactions are surely where the real action (haha) is, and even if it’s a tiny fraction of total views in the country/continent/globe/galaxy, it still matters for how it drives American thinking, or just maintains it. “Widely viewed” content might leak across borders, but that’s not quite the same thing. (Roose thinks it’s odd too, unique link to this extract
A CGI replica of Nvidia’s CEO delivered his keynote and no one knew • Input Mag
Tom Maxwell:
»
Most people hate giving presentations, but thankfully someday you might be able to give one without actually delivering it yourself at all. Nvidia, the maker of popular graphics cards, revealed yesterday that parts of a keynote speech made by its CEO were actually computer-generated animation — an entire virtual replica of Jensen Huang and his kitchen in the background.
The speech happened in April, and only about 14 seconds of the nearly two-hour presentation were animated. But part of the presentation showed Huang magically disappear and his kitchen explode, which made viewers wonder what exactly was real or rendered. It’s hard to actually identify the fake portion, however, which is the most impressive part.
Granted, creating the rendered version of Huang involved a lot of work. Using a truck full of DSLR cameras, a full face and body scan was captured to create a 3D model, and then artificial intelligence was trained to mimic his gestures and expressions. Nvidia says it also applied some other “AI magic” to make his clone look realistic. Even still, it’s quite an impressive feat that one can watch the presentation and still not be sure what parts of actual, recorded video, and which are fake.
«
Impressive, though they must have known just how good it would look. It’s not a risk if you can prepare sufficiently.
unique link to this extract
Yes, yes, it’s the plug for
Social Warming, my latest book. Social media, outrage, amplification, and how they interact and make us behave worse and worse.
Leaps, bounds, and backflips • Boston Dynamics
Calvin Hennick on the latest video showing two humanoid Boston Dynamics robots doing kinda-sorta parkour:
»
Simulation is an essential development tool for the Atlas controls team, both for evaluating new behaviors prior to robot testing and for ensuring that new software changes don’t negatively impact existing capabilities. But there’s still no replacement for hardware testing, particularly in performance-limiting motions like vaulting.
About that vault: unlike high-flipping gymnastics vaults, a parkour vault is a slightly less flashy method designed to get a runner over a low wall or obstacle – in this case the balance beam, only a few feet high. Atlas places its arm on the beam and then hoists its body over the structure. For many humans, this sort of vault would be relatively easy (especially in comparison to a backflip), but for the Atlas team, it represented a formidable new challenge.
“If you or I were to vault over a barrier, we would take advantage of certain properties of our bodies that would not translate to the robot,” Kuindersma notes. “For example, the robot has no spine or shoulder blades, so it doesn’t have the same range of motion that you or I do. The robot also has a heavy torso and comparatively weak arm joints. Extending our tools to help us find solutions that worked within these constraints was what made the vault an interesting challenge.”
During filming, Atlas gets the vault right about half of the time. (A natural consequence of pushing robots to their limit is that, sometimes, those limits are met.) On the other runs, Atlas makes it over the barrier, but loses its balance and falls backward, and the engineers look to the logs to see if they can find opportunities for on-the-fly adjustments.
“There are a lot of pretty exciting behaviors here, and some of them are not totally reliable yet,” says Ben Stephens, the Atlas controls lead. “Every behavior here has a small chance of failure. It’s almost 90 seconds of continuous jumping, jogging, turning, vaulting, and flipping, so those probabilities add up.”
Stephens adds that this is the first time Boston Dynamics has filmed two robots performing parkour together. “We had never actually done the two robots together until two weeks ago,” he says on the day the routine is being filmed. “We’re in a place now where it should work. We think we’ve caught all of the major failures, and now it’s just down to those small probabilities.”
«
Doesn’t quite inspire the feeling that these are really autonomous robots, though.
unique link to this extract
DDoSecrets is the new WikiLeaks • The New Republic
Jacob Silverman:
»
Whereas WikiLeaks cultivated an anti-imperialist mystique centered on the cultish figure of Assange, DDoSecrets professes something more modest: an unvarnished commitment to providing information useful to journalists and concerned citizens. As the DDoSecrets website puts it, data must fulfill two criteria: “Is it in the public interest?” and “Can a prima facie case be made for the veracity of its contents?” If it passes that test—and the group, which now has approximately 10 members along with an advisory board and volunteer contributors, decides collectively that they can protect their sources—then they publish the archive, sometimes as an easily downloadable torrent, other times through its slightly more difficult to reach onion site, which requires using the Tor browser. While many archives are published for a wide audience, others are withheld and only offered to journalists upon request; and in some cases, the organization will write about data it receives without publishing its contents.
At its best, the work of DDoSecrets reveals the limits of official transparency, of authorized government leaks and incrementalist beat reporting and FOIA requests that yield pages of useless redactions. Nowhere is this more visible than with BlueLeaks. “Reading the unredacted, hacked documents gives a very different picture than the selections you get from an open records officer,” said Brendan McQuade, author of Pacifying the Homeland, a book about the modern surveillance state. Based on BlueLeaks information [269GB of data about US police lawbreaking and surveillance overreach], he wrote articles that exposed police malfeasance and brought attention to a federal whistleblower suit against the Maine Information and Analysis Center, or MIAC. Maine’s state house later voted to close the site (although the bill never cleared the Senate). To McQuade, and to the members of DDoSecrets, hacked data provides what official channels cannot: truth and the potential for accountability.
«
Twitter bans DDoSecrets, and you can’t post a URL to it on there without it being removed, because it uses hacked material. (How does that not apply to Wikileaks, exactly?) Lucky for the Taliban they didn’t take over Afghanistan by hacking anything, I guess.
unique link to this extract
Galaxy Watch 4 review: welcome to Samsung’s garden • The Verge
Dieter Bohn reviews the offspring of Google and Samsung’s semi-forced Wear OS marriage:
»
I really enjoyed using the Galaxy Watch 4. It has been a genuine pleasure to have a competent and capable smartwatch paired to my Android phone — one that doesn’t have any show-stopping problems.
But the reason I had that nice experience was because I was using the Galaxy Watch 4 with a Samsung phone. If you’re a Samsung user, the Galaxy Watch 4 is an excellent smartwatch. If you’re not, the Galaxy Watch 4 all but forces you into Samsung’s ecosystem. Samsung’s ecosystem is better than it often gets credit for, but it’s limiting. Just as the Apple Watch keeps people on the iPhone, Samsung’s watch will keep people on Samsung phones (or at least get them to install Samsung software and use Samsung services on their Android phones).
If you’d like to use a Wear OS 3 smartwatch that isn’t tied to Samsung, I wish I knew what to tell you. There are no Wear OS 3 smartwatches from other manufacturers on the horizon. After so many years of waiting for a good smartwatch for Android users, it’s finally here — but only for some of us.
«
Smartwatches are yet another illustration of how the modular approach (different companies making the chips, hardware, software, OS) is not always the path to dominance in technology. (See also: MP3 players, tablets.)
unique link to this extract
Eco-friendly vehicles offer quieter, cleaner safaris in Kenyan reserve • Reuters
Monicah Mwangi:
»
In Kenya’s Maasai Mara National Reserve, the Toyota 4×4 Landcruiser of tour guide and driver Sylvester Mukenye glides silently past a herd of grazing elephants, then past a pride of lions lying in the grass.
The animals are completely unperturbed by the proximity of the vehicle because its diesel engine has been replaced by an electric one that eliminates the rumbling noise and, just as importantly, reduces the emission of diesel fumes.
“If you drive here silently, you will of course get much closer to animals, especially the elephants that we are next to right now, because there are no vibrations on the ground and there are no fumes that they get the smell from like in other cars,” Mukenye said.
His vehicle was converted by Opibus, a Nairobi-based Kenyan-Swedish company founded in 2017. It is, for now, the only company in Kenya that converts off-road safari vehicles from diesel and petrol to electric power.
Off-road vehicles are a common sight in Maasai Mara, world-famous for the annual wildebeest migration but these are the first in the usually carbon-heavy business of safari tours to be entirely powered by electric batteries.
Wanjiru Kamau, an electrical engineer at Opibus, said the company had so far converted 10 vehicles used in Kenyan game parks, including three in the Maasai Mara. As well as being more environmentally friendly than diesel engines, the electric motors cut operating costs by half, she added.
“In Kenya our fuel prices are always rising… Why not save on that?” she told Reuters at the Opibus workshop, where assembled vehicles were in various stages of electrification.
«
And of course you’d use solar power in Kenya. The accompanying video is a great watch as well.
unique link to this extract
Evidence of fraud in an influential field experiment about dishonesty • Data Colada
An anonymous group of three researchers – “Uri, Joe and Leif”:
»
In 2012, Shu, Mazar, Gino, Ariely, and Bazerman published a three-study paper in PNAS reporting that dishonesty can be reduced by asking people to sign a statement of honest intent before providing information (i.e., at the top of a document) rather than after providing information (i.e., at the bottom of a document). In 2020, Kristal, Whillans, and the five original authors published a follow-up in PNAS entitled, “Signing at the beginning versus at the end does not decrease dishonesty”. They reported six studies that failed to replicate the two original lab studies, including one attempt at a direct replication and five attempts at conceptual replications.
Our focus here is on Study 3 in the 2012 paper, a field experiment (N = 13,488) conducted by an auto insurance company in the southeastern United States under the supervision of the fourth author. Customers were asked to report the current odometer reading of up to four cars covered by their policy. They were randomly assigned to sign a statement indicating, “I promise that the information I am providing is true” either at the top or bottom of the form. Customers assigned to the ‘sign-at-the-top’ condition reported driving 2,400 more miles (10.3%) than those assigned to the ‘sign-at-the-bottom’ condition.
The authors of the 2020 paper did not attempt to replicate that field experiment, but they did discover an anomaly in the data: a large difference in baseline odometer readings across conditions, even though those readings were collected long before – many months if not years before – participants were assigned to condition. The condition difference before random assignment (~15,000 miles) was much larger than the analyzed difference after random assignment (~2,400 miles):
In trying to understand this, the authors of the 2020 paper speculated that perhaps “the randomization failed (or may have even failed to occur as instructed) in that study” (p. 7104).
On its own, that is an interesting and important observation. But our story really starts from here, thanks to the authors of the 2020 paper, who posted the data of their replication attempts and the data from the original 2012 paper (.htm). A team of anonymous researchers downloaded it, and discovered that this field experiment suffers from a much bigger problem than a randomization failure: there is very strong evidence that the data were fabricated.
«
Paper about dishonesty may be dishonest. No wonder it couldn’t be replicated.
unique link to this extract
[P] AppleNeuralHash2ONNX: Reverse-Engineered Apple NeuralHash, in ONNX and Python • Reddit MachineLearning
Asuhariet Ygvar:
»
As you may already know Apple is going to implement NeuralHash algorithm for on-device CSAM detection soon. Believe it or not, this algorithm already exists as early as iOS 14.3, hidden under obfuscated class names. After some digging and reverse engineering on the hidden APIs I managed to export its model (which is MobileNetV3) to ONNX and rebuild the whole NeuralHash algorithm in Python. You can now try NeuralHash even on Linux!
Source code: https://github.com/AsuharietYgvar/AppleNeuralHash2ONNX
No pre-exported model file will be provided here for obvious reasons. But it’s very easy to export one yourself following the guide I included with the repo above. You don’t even need any Apple devices to do it.
Early tests show that it can tolerate image resizing and compression, but not cropping or rotations.
Hope this will help us understand NeuralHash algorithm better and know its potential issues before it’s enabled on all iOS devices.«
To be clear, this is the system that detects CSAM in your photo library before upload. The writer does explain why s/he is confident the system as recreated is what’s there. Apparently there are 200 “layers” of neural network in the hashing system, though I don’t know if that’s a lot or a little (I suspect “a lot”).
The discovery opens up lots of possibilities, such as (one person suggested) training a generative adversarial network (GAN) to create images that trip it yet aren’t CSAM. (One collision of two non-CSAM photos has already been found.) But of course Apple will just check them manually, and if there are too many innocent collisions it will raise the threshold above 30 photos.
unique link to this extract
Brazilian stripe adventure • Dave Birch’s blog
Dave Birch on Mastercard’s announcement that it will stop issuing cards with magnetic stripes from 2024:
»
A few years ago, and already a few years after I had been enjoying the fact that Kazakhstan was migrating to chip and pin when Kansas wasn’t, I had to go to São Paulo for a few days to work with some of the Brazilian banks. I can’t remember why, exactly ,but I think it was something to do with mobile payments. Anyway, when it came time for me to leave I found a taxi and set off for the airport.
While pottering along the freeway I remembered that I didn’t have any cash with me, because I never do, and so I wanted to know if the taxi could take payment by card. I took out my wallet and gestured at a credit card and looked quizzically at the driver. The driver signalled and turned off of the freeway onto some side roads. After a few minutes of driving through a retail area, which is mainly shoe shops as I recall, we turned off again onto some smaller roads and went into a distinctly shady part of town.
At this point I began to panic slightly.
Cursing my stupidity for waving around a wallet full of cards and naturally assuming that I was about to be robbed, I began to calmly assemble my tactics. I figured that so long as I could retain my passport then things would be okay. After all, credit cards could be replaced by the banks (as indeed they often were at though at that time) and the computer belonged to the company not me, so whatever. I surreptitiously removed my passport from my jacket pocket and hoping that the driver would not notice, slid it down my leg and into one of my socks where I hoped it would remain throughout my impending ordeal.
The car pulled up at a shack that looked for all the world like a bandit headquarters from Mad Max and the driver shouted something to the unseen denizens. I thought for a moment about trying to make a run for it but realised I wouldn’t really get very far and would likely only inflame the situation, so I stayed put to await the inevitable. Sure enough, a young man dressed in jeans and some sort of football shirt came out from behind the shack and jogged towards the car with something metallic in his hand.
He reached the car and pulled open the door and thrust toward me, glinting in the sunlight… a chip and pin terminal.
I put the card in, punched in my pin, waited for my receipt and continued to the airport.
«
Americans have no idea how backward their financial systems are. He also had this post from 2014 asking why his cards have mag strips at all.
unique link to this extract
How extreme weather is hitting airlines • Financial Times
Claire Bushey and Philip Georgiadis:
»
This month, storms forced the cancellation of more than 300 flights at both Chicago’s O’Hare airport and Dallas/Fort Worth airport in Texas. In July, eight flights in Denver were cancelled and another 300 delayed due to smoke from forest fires burning in the US Pacific Northwest. Extreme heat affected take-offs in Las Vegas and Colorado earlier this summer.
The disruptions are in line with a trend: weather-related flight cancellations and delays have increased over the past two decades in the US and Europe, regulatory data shows. While it is difficult to link any individual storm or heatwave to climate change, scientific studies have found they will become more frequent or intense as Earth grows warmer.
The International Civil Aviation Organization, the UN standard-setting body, found in a 2019 poll of member states that three-quarters of respondents said the airline industry already was experiencing some impact from climate change.
“It is something that is absolutely on our minds, as far as how we’re going to be able to continue to run the flight schedule, especially with the growth that we have planned for the future,” said David Kensick, managing director of global operations at United Airlines. “With climate change, we are seeing some of that weather that’s hard to predict, so we need to be better at dealing with it.”
Airlines contribute about 2% of global carbon-dioxide emissions globally, though counting other substances spewed from aircraft, some studies indicate their climate impact is bigger.
The potential impacts of climate change on the industry are far-reaching. In the short term, intense weather conditions present an operational headache. Forced flight diversions and cancellations add costs to an industry that haemorrhaged billions of dollars during the pandemic.
In the longer term, airlines believe changing wind patterns will alter flight routes and fuel consumption. It will probably take longer to fly from Europe to the US as the jet stream above the north Atlantic changes, for example.
“Aviation will be a victim of climate change as well as, in many people’s eyes, a villain,” said Paul Williams, professor of atmospheric science at the University of Reading in the UK.
«
Once again, the thing of not seeing the problem if your salary depends on not seeing it.
unique link to this extract
Errata, corrigenda and ai no corrida: none notified
The Overspill: when there's more that I want to say 