A new Google project uses machine learning to try to optimise traffic light timings to improve vehicle flow in a growing number of cities. CC-licensed photo by Ian Sane 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. Go, go, go! I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.
Did you miss the post on the Social Warming Substack last Friday? It looked at what sort of difference the alterations to news item display on Twitter might make to clickthroughs – and what engagement is about.
Why the internet isn’t fun anymore • The New Yorker
Kyle Chayka:
»
The social-media Web as we knew it, a place where we consumed the posts of our fellow-humans and posted in return, appears to be over. The precipitous decline of X is the bellwether for a new era of the Internet that simply feels less fun than it used to be. Remember having fun online? It meant stumbling onto a Web site you’d never imagined existed, receiving a meme you hadn’t already seen regurgitated a dozen times, and maybe even playing a little video game in your browser. These experiences don’t seem as readily available now as they were a decade ago. In large part, this is because a handful of giant social networks have taken over the open space of the Internet, centralizing and homogenizing our experiences through their own opaque and shifting content-sorting systems. When those platforms decay, as Twitter has under Elon Musk, there is no other comparable platform in the ecosystem to replace them. A few alternative sites, including Bluesky and Discord, have sought to absorb disaffected Twitter users. But like sproutlings on the rain-forest floor, blocked by the canopy, online spaces that offer fresh experiences lack much room to grow.
…The Internet today feels emptier, like an echoing hallway, even as it is filled with more content than ever. It also feels less casually informative. Twitter in its heyday was a source of real-time information, the first place to catch wind of developments that only later were reported in the press. Blog posts and TV news channels aggregated tweets to demonstrate prevailing cultural trends or debates. Today, they do the same with TikTok posts—see the many local-news reports of dangerous and possibly fake “TikTok trends”—but the TikTok feed actively dampens news and political content, in part because its parent company is beholden to the Chinese government’s censorship policies. Instead, the app pushes us to scroll through another dozen videos of cooking demonstrations or funny animals. In the guise of fostering social community and user-generated creativity, it impedes direct interaction and discovery.
«
The Israel-Hamas war shows just how broken social media has become • The Atlantic
Charlie Warzel:
»
As my colleague Kaitlyn Tiffany wrote last year, the history of war is a history of media. The Gulf War demonstrated the power of CNN and the 24/7 cable-news format, foreshadowing the way infotainment would permeate politics and culture for the next 20 years. A series of contentious election cycles from 2008 to 2020, as well as the Arab Spring, the Syrian civil war, and the rise of the Islamic State, showed how social-media platforms democratized punditry and journalism, for better and worse. Commentators were quick to dub Russia’s invasion of Ukraine the “first TikTok war,” as the internet filled with videos from Ukrainians documenting the horrors of war in profoundly personal, often surreal ways.
If such conflicts are lenses through which we can understand an information environment, then one must surmise that, at present, our information environment is broken. It relies on badly maintained social-media infrastructure and is presided over by billionaires who have given up on the premise that their platforms should inform users. During the first days of the Israel-Hamas war, X owner Elon Musk himself has interacted with doctored videos published to his platform. He has also explicitly endorsed accounts that are known to share false information and express vile anti-Semitism.
In an interview with The New York Times, a Hamas official said that the organization has been using the lack of moderation on X to post violent, graphic videos on the platform to terrorize Israeli citizens.
Meanwhile, Adam Mosseri, the head of Instagram and the unofficial lead on the company’s Twitter clone, Threads, has received requests from journalists, academics, and news junkies to make his product more useful for following the war. He has responded by saying that his team won’t “amplify” news media on the platform
«
Tesla owner in Israel escapes terrorists in Model 3 despite blown tires, dozens of bullet holes • Teslarati
Simon Alvarez:
»
During the beginning of Hamas’ attack, a Tesla owner from Mefalsim, a kibbutz in Southern Israel near the Gaza Strip, was called in with the community’s alert squad. As he was driving to the assembly point, he encountered a vehicle loaded with Hamas terrorists. Photos and videos of the Model 3 Performance after its run-in with the terrorists hinted at the shocking events that transpired.
Speaking from Sheba Hospital, where he is recovering from a series of surgeries due to the attack, the Tesla owner told Israeli publication Walla about his encounter. According to the Model 3 owner, the terrorists proceeded to spray the Model 3 with bullets, shooting at the front in an attempt to hit the engine. The terrorists also shot at the Model 3’s rear, seemingly in an attempt to ignite the fuel tank. Fortunately, the Model 3 Performance had neither an engine nor a fuel tank.
“The terrorists recognized me from a distance of 10 meters. In addition to their Kalashnikovs, they had a machine gun in the battle that fired bullets of a larger diameter. They didn’t realize it was an electric car, so they shot at the front, hoping to hit the engine that wasn’t there, and at the back, attempting to ignite the non-existent fuel. They shot my tires. I pressed the gas, and they started chasing me,” the Tesla owner noted.
At this point, it was a matter of survival. Thankfully, the Model 3 Performance is a very quick car, and the Tesla owner was able to gain some distance from his attackers. The Model 3 owner noted that his car’s acceleration ultimately allowed him to get away, and the Tesla’s safety systems allowed him to drive to a hospital at speed, even with blown-out tires.
«
The photos are dramatic (and the driver was badly wounded) but it certainly shows one benefit of EVs. He escaped his would-be killers by driving at up to 180km/h (110mph) with the tyres disintegrating.
unique link to this extract
X fined $610,500 in Australia first for failing to crack down on child sexual abuse material • The Guardian
Josh Taylor:
»
X, the company formerly known as Twitter, has become the first online platform to be issued with a $610,500 fine under Australia’s Online Safety Act for its failure to meet basic online safety expectations.
X has 28 days to either pay the fine, issued by the e-safety commissioner, Julie Inman Grant, or provide responses to questions X ignored from the commissioner on its work to crack down on child sexual abuse material on the platform.
The legal notices were issued to X, Google, TikTok, Twitch and Discord in February following the first round of notices sent to Apple, Meta, Microsoft, Snap and Omegle last year.
In a report on the tech companies’ responses to the notices, released on Monday, the commissioner found many of the largest services were failing to adequately detect, remove and prevent child abuse material.
“Frankly, I was surprised at how hard it was to extract precise and accurate information, and frankly, surprised that some companies that should have much more sophisticated and mature systems and resources didn’t seem willing or able to be able to provide that information that had been provided by other companies,” Inman Grant told Guardian Australia. “Or in the case of Twitter, to leave things totally blank, to obfuscate [from providing] inaccurate information.”
The commissioner also found that X and Google did not comply with the notices, with Google giving generic responses to some specific questions, while some questions to X went entirely unanswered.
Google has been given a formal warning, while X was given an infringement notice.
«
This is just the beginning. Twitter is in line for many more, repeated fines all over the place. The EU and the US’s FTC in particular.
unique link to this extract
The same lab that cloned Dolly the sheep has used gene editing to create chickens resistant to avian flu • EL PAÍS English
Daniel Mediavilla:
»
The University of Edinburgh’s Roslin Institute — the animal research center where Dolly the sheep was created — used gene editing to breed chickens that resist infection by the avian flu. A deadly virus for birds that causes great economic losses around the world and can, in some cases, infect and kill humans, the disease has proved difficult for vaccines because the proteins on its surface that are recognized by the immune system change rapidly. A group of British researchers has tested the potential of modifying small sections of chicken DNA to prevent influenza infection, albeit only partially. They published their results today in the journal Nature Communications.
Influenza A needs a protein in chicken cells, ANP32A, to replicate. The team of scientists, led by Mike McGrew, a University of Edinburgh researcher, used the CRISPR editing technique to modify the gene that produces the protein in the chickens’ germ cells, which would enable the birds to pass down the change to their offspring. In this way, animals were created that hardly became infected with influenza when exposed to other infected birds (“9 out of 10 remained uninfected,” according to the study), and they did not subsequently infect other chickens. In a later test, when inoculated with a dose a thousand times higher, five out of ten became infected.
«
This is quite a feat. Avian flu has had a dramatic effect on farming in Europe.
unique link to this extract
Young people turn to social media for financial guidance • Deloitte UK
»
More young people are turning to social media rather than their bank for financial guidance, according to new research from Deloitte.
A survey of over 2,500 UK consumers, conducted in August 2023, found that 25% of 18-24 year olds use social media when searching for financial guidance and advice. It also found that 20% of this age group have invested money based on social media recommendations, with almost half of these (48%) having invested between £100-£500 and 16% invested over £1000, in their lifetime. 21% invested specifically in cryptocurrency based on social media guidance. Yet, 33% of the same age group are not confident enough in their financial knowledge to take out investment products at all.
The survey – which aims to understand the impact of the rise in the cost of living on banking and insurance customers – reveals that the majority of consumers across all age groups are turning to alternative sources for financial guidance and advice. Only 16% of respondents in the banking survey said they would seek guidance from their bank, with respondents preferring to seek it elsewhere, such as from their friends and family (34%) and the MoneySavingExpert (25%). The main reasons cited for this was that they were either unsure of what supportive services their bank has to offer, or too embarrassed to seek out support.
«
Perhaps this, rather than avocado toast, is the explanation for a generation’s inability to buy a house. Well, for why so many were scammed by crypto.
unique link to this extract
Control Panel for Twitter • Chrome Web Store
Jonathan Buchanan:
»
Control Panel for Twitter is a [Chrome] browser extension for Twitter – its main goals are to reduce the amount of algorithmic content you see, give you more control over your timeline and make the UI less distracting.
By default,
• your Home timeline will be the reverse-chronological “Following” timeline, which will only contain tweets and quote tweets from the people you follow.• The algorithmic “For you” timeline will be hidden, and Control Panel for Twitter will keep you on “Following” if Twitter tries to automatically take you off it.
• Everything is configurable, so start by opening the options popup and customizing to your liking. Changes you make in the options will be applied immediately.
«
Also gets rid of ALL of the paid-for Twitter Blue accounts. Great for Chrome users.
unique link to this extract
Google’s AI stoplight program is now calming traffic in a dozen cities worldwide • Engadget
Andrew Tarantola:
»
It’s been two years since Google first debuted Project Green Light, a novel means of addressing the street-level pollution caused by vehicles idling at stop lights. At its Sustainability ‘23 event on Tuesday, the company discussed some of the early findings from that program and announced another wave of expansions for it.
Green Light uses machine learning systems to comb through Maps data to calculate the amount of traffic congestion present at a given light, as well as the average wait times of vehicles stopped there. That information is then used to train AI models that can autonomously optimize the traffic timing at that intersection, reducing idle times as well as the amount of braking and accelerating vehicles have to do there. It’s all part of Google’s goal to help its partners collectively reduce their carbon emissions by a gigaton by 2030.
When the program was first announced in 2021, it had only been pilot tested in four intersections in Israel in partnership with the Israel National Roads Company but Google had reportedly observed a “10 to 20% reduction in fuel and intersection delay time” during those tests. The pilot program has grown since then, spreading to a dozen partner cities around the world, including Rio de Janeiro in Brazil; Manchester in England and Jakarta in Indonesia.
“Today we’re happy to share that… we plan to scale to more cities in 2024,” Yael Maguire, Google VP of Geo Sustainability, told reporters during a pre-brief event last week. “Early numbers indicate a potential for us to see a 30% reduction in stops. We believe green light is unique because it is more scalable and cost effective for cities than alternative options.”
«
A new tool helps artists thwart AI—with a middle finger • WIRED
Kate Knibbs:
»
It’s helpful to know how image generators are trained to understand exactly how Kudurru works. Most of these generators find their training data by “scraping” the internet. Scrapers use software that collects data in bulk from across the web, from platforms like DeviantArt and professional libraries like Getty Images to individual artists’ websites. One of the most popular and most commonly used roadmaps to decide what to scrape is the dataset LAION-5B, which lists the URLs to billions of images. When an AI company uses a dataset like LAION-5B to scrape images, it has to download those images from the URL links. That’s where Kudurru finds its opening.
According to Spawning cofounder Jordan Meyer, during internal testing, Kudurru was able to briefly stymie a substantial amount of scraping activity. “For about two hours in July, we stopped everyone who was in the process of downloading the LAION-5B dataset,” Meyer says.
To identify the scrapers, Spawning operates a honeypot-like “defense network” of more than 1,000 websites, each hosting images that groups using LAION-5B would scrape to train a generative AI model. These websites collect data on the IP addresses attempting to scrape images; Spawning can often identify the groups doing the scraping and the regions with the most overall scraping activity. (China is currently in the lead.)
“We’re developing what is basically a blacklist,” Spawning cofounder Patrick Hoepner says. Spawning, also the company behind Have I Been Trained?, a site that lets creators see if AI has scraped their work, updates this blacklist in real time, based on the behavior of the IP addresses it tracks.
Kudurru gives artists two options to disrupt scraping. First, they can simply block the blacklisted IP addresses. Second, to take things a step further, they can also choose to sabotage or “poison” the scrapers’ efforts by sending back a different image than the one requested. Spawning gives users the option to choose what images they send back, although it does have some suggestions. “It could just be a middle finger over and over again,” Meyer says.
«
Inside the deadly instant loan app scam that blackmails with nudes • BBC News
Poonam Agarwal:
»
The business model is brutal but simple.
There are many apps that promise hassle-free loans in minutes. Not all of them are predatory. But many – once downloaded – harvest your contacts, photos and ID cards, and use that information later to extort you. When customers don’t repay on time – and sometimes even when they do – they share this information with a call centre where young agents of the gig economy, armed with laptops and phones are trained to harass and humiliate people into repayment.
At the end of 2021, Bhoomi had borrowed about 47,000 rupees ($565; £463) from several loan apps while she waited for some work expenses to come through. The money arrived almost immediately but with a big chunk deducted in charges. Seven days later she was due to repay but her expenses still hadn’t been paid, so she borrowed from another app and then another. The debt and interest spiralled until she owed about two million rupees ($24,000; £19,655).
Soon the recovery agents started calling. They quickly turned nasty, slamming Bhoomi with insults and abuse. Even when she had paid, they claimed she was lying. They called up to 200 times a day. They knew where she lived, they said, and sent her pictures of a dead body as a warning.
As the abuse escalated they threatened to message all of the 486 contacts in her phone telling them she was a thief and a whore. When they threatened to tarnish her daughter’s reputation too, Bhoomi could no longer sleep.
She borrowed from friends, family and more and more apps – 69 in total. At night, she prayed the morning would never come. But without fail at 07:00, her phone would start pinging and buzzing incessantly.
Eventually, Bhoomi had managed to pay back all of the money, but one app in particular – Asan Loan – wouldn’t stop calling. Exhausted, she couldn’t concentrate at work and started having panic attacks.
One day a colleague called her over to his desk and showed her something on his phone – a naked, pornographic picture of her.The photo had been crudely photoshopped, Bhoomi’s head stuck on someone else’s body, but it filled her with disgust and shame. She collapsed by her colleague’s desk. It had been sent by Asan Loan to every contact in her phone book. That was when Bhoomi thought of killing herself.
«
At least 60 people in India have killed themselves after being targeted in this scam.
unique link to this extract
| • Why do social networks drive us a little mad? • Why does angry content seem to dominate what we see? • How much of a role do algorithms play in affecting what we see and do online? • What can we do about it? • Did Facebook have any inkling of what was coming in Myanmar in 2016? Read Social Warming, my latest book, and find answers – and more. |
Errata, corrigenda and ai no corrida: none notified
The Overspill: when there's more that I want to say 