Great spotted eagles changed their migration routes to avoid Ukraine once the conflict began there in 2022. CC-licensed photo by Bernard DUPONT 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.
There’s another post coming this week at the Social Warming Substack on Friday at 0845 UK time. Free signup.
A selection of 10 links for you. Like Steve Miller? I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.
How the new Microsoft Recall feature fundamentally undermines Windows security • DoublePulsar
Kevin Beaumont:
»
On Monday, Microsoft CEO Satya Nadella sat down with the media to introduce a new feature called Recall, as part of their Copilot+ PCs. It takes screenshots of what you’re doing on constantly, by design:
The idea is it allows you to rewind back in time at the click of a button to see what you were doing at, say, 11pm two months ago. It also classifies almost everything you’re doing, seeing and typing. This is instantly searchable.
Microsoft’s store page for the PCs points out “It will not hide information such as passwords or financial account numbers [..]” For example, if you log into online banking, your information around account numbers, balances, purchases etc will enter Recall’s database.
This fundamentally changes the relationship with you and your Microsoft Windows computer. It also introduces real risk to you, the customer. Let’s break down what is happening.
You may look at this and think ‘surely there’s some safety guardrails’, and there are — in the video above you’ll see Satya point out the processing and data storage is done locally on the device. In the FAQ they point out there’s some circumstances where data won’t be recorded, for example when password’s aren’t visible on screen.
That’s great. It’s also not nearly enough.
If you look at what has happened historically with infostealer malware — malicious software snuck onto PCs — it has pivoted to automatically steal browser passwords stored locally. In other words, if a malicious threat actor gains access to a system, they already steal important databases stored locally.
«
You can turn it off, but of course then.. you’ve turned it off. This is going to be a hell of a target for malware.
unique link to this extract
The tragic downfall of the internet’s art gallery • Slate
Nitish Pahwa:
»
On March 27, a large group of artists and creators from across the web noticed the frightening extent to which a once-beloved, highly influential community platform of theirs had, like so many others, fallen prey to the artificial intelligence juggernauts plundering the internet.
As VFX animator Romain Revert (Minions, The Lorax) pointed out on X, the bots had come for his old home base of DeviantArt. Its social accounts were promoting “top sellers” on the platform, with usernames like “Isaris-AI” and “Mikonotai,” who reportedly made tens of thousands of dollars through bulk sales of autogenerated, dead-eyed 3D avatars. The sales weren’t exactly legit—an online artist known as WyerframeZ looked at those users’ followers and found pages of profiles with repeated names, overlapping biographies and account-creation dates, and zero creations of their own, making it apparent that various bots were involved in these “purchases.”
It’s not unlikely, as WyerframeZ surmised, that someone constructed a low-effort bot network that could hold up a self-perpetuating money-embezzlement scheme: Generate a bunch of free images and accounts, have them buy and boost one another in perpetuity, inflate metrics so that the “art” gets boosted by DeviantArt and reaches real humans, then watch the money pile up from DeviantArt revenue-sharing programs. Rinse, repeat.
After Revert declared this bot-on-bot fest to be “the downfall of DeviantArt,” myriad other artists and longtime users of the platform chimed in to share in the outrage that these artificial accounts were monopolizing DeviantArt’s promotional and revenue apparatuses. Several mentioned that they’d abandoned their DeviantArt accounts—all appearing to prove his dramatic point.
«
Stalk my date: I know where you were last night • Bustle
Kate Lindsay:
»
Location-monitoring is all but expected in relationships nowadays, in part because it comes in so many forms — from Life360 to Snap Map. Whether they mean to or not, users are constantly signaling their availability or where they are in real time. Instagram beams out a bright green dot next to my username whenever I’m scrolling; Facebook Messenger allows me to toggle sharing my whereabouts. It’s so easy, people may not even know when and where they’re doing it. Just ask 31-year-old Carlotta*, whose ex seemed to have no idea he had turned location services on for Facebook Messenger while they were dating.
“It was definitely by accident,” she told me. “He would message me, ‘Just leaving my house for work. What are you doing tonight?’ I would see in his location services that he was somewhere completely different.” She used this information to confront him, and he admitted to sleeping with someone else.
For those with an iPhone, one location-sharing feature, Find My Friends, dominates. Apple launched its standalone app in 2011, and by 2015, it came automatically with new iPhones. Since it merged with Find My iPhone and Find My Mac in 2019, Find My has been a one-stop shop for all users’ surveillance needs. Even when I’m not actively checking in on my friends and family, my iPhone is: I’ll reliably get an alert when my boyfriend (or, really, his AirPods) have been near me for a while.
Just last week, my friend Selina Scharr, a 31-year-old from Brooklyn, dropped her dog off at the vet for a stressful surgery. After an emotional goodbye, she left, only to receive an alert from the AirTag she keeps on her dog’s collar. Basil, it warned her, had been “left behind.”
“Jeez, Apple, way to make me cry this morning,” she texted our group chat. (Basil is recovering happily at home.)
This technology has changed everything. I no longer have to ask my friend for their ETA or wonder if my parents are available for a phone call. I can just check in on their real-time location, a capability previously associated with intelligence agencies and covert criminals. What is, for all intents and purposes, stalking, has become a casual part of everyday familial, romantic, and platonic relationships — and turned many with formerly healthy boundaries into lurkers in the process.
«
Everyone in Hollywood is using AI, but “they are scared to admit it” • Hollywood Reporter
Winston Cho and Scott Roxborough:
»
For horror fans, Late Night With the Devil marked one of the year’s most anticipated releases. Embracing an analog film filter, the found-footage flick starring David Dastmalchian reaped praise for its top-notch production design by leaning into a ’70s-era grindhouse aesthetic reminiscent of Dawn of the Dead or Death Race 2000. Following a late-night talk show host airing a Halloween special in 1977, it had all the makings of a cult hit.
But the movie may be remembered more for the controversy surrounding its use of cutaway graphics created by generative artificial intelligence tools. One image of a dancing skeleton in particular incensed some theatergoers. Leading up to its theatrical debut in March, it faced the prospect of a boycott, though that never materialized.
The movie’s directors Cameron and Colin Cairnes defended the AI usage, explaining the art was touched up by human hands. In a statement, they said, “We experimented with AI for three still images which we edited further and ultimately appear as very brief interstitials in the film.”
Less than a month later, five images suspected to be generated by AI teasing postapocalyptic scenes in A24’s Civil War sparked similar outrage by a segment of fans. There were a few telltale signs that the graphics were AI-created in landmark accuracy and consistency blunders: The two Chicago Marina Towers buildings in one poster are on opposite sides of the river; in another, a shot of wreckage shows a car with three doors.
In response, a reader on A24’s Instagram post wrote that the backlash to Late Night was “more than enough to make transparently clear to everyone: WE DO NOT WANT THIS.”
But in the entertainment industry, the Pandora’s box of AI has likely already been unleashed. Behind closed doors, most corners of production, from writers’ rooms to VFX departments, have embraced generative AI tools.
«
Those Civil War posters (in the linked article) do look weird, and AI-adjacent – especially the car with three doors on one side. They aren’t scenes from the (excellent) film itself.
unique link to this extract
Four more cats die of H5N1 bird flu in the US • BNO News
»
Four more cats have died of H5N1 bird flu in the United States, including two pets in South Dakota with no links to poultry or dairy cows, according to state and federal officials. At least 14 cats have recently died of bird flu.
Of the newly reported cases, two were domestic cats which died at a property in Campbell County in South Dakota, according to a state official and the U.S. Department of Agriculture (USDA).
Beth Thompson, the state’s veterinarian, said there was no livestock on the property where the pets died. “No other details regarding how the cats were infected are known at this time,” Thompson told BNO News.
Two other cases were recently reported in Michigan, one in Isabella County and the other in Ionia County. Both cases involved barn cats on commercial dairy farms where cows were also infected with H5N1.
At the property in Ionia County, two Virginia opossums were also infected with the bird flu virus.
“Cats are particularly susceptible to H5N1 2.3.4.4b viruses and the majority of sick cats have been reported at or near affected poultry facilities or dairies,” Shilo Weir, a spokesperson for USDA’s Animal and Plant Health Inspection Service, told BNO News.
…In the United States, at least 27 cats have now been infected with H5N1 bird flu, including the 14 cases reported in recent weeks. The other 13 happened last year in connection with infected poultry or wild birds.
«
Watching brief. (Also: birds, cows, opossums, cats. It’s as if viruses can cross species barriers.)
unique link to this extract
Why we’re unhappiest in our late 40s • Greater Good
Jill Suttie:
»
According to new research, the United States isn’t the only place where midlife is so hard. Studying 132 countries around the world, labor economist and researcher David Blanchflower found strong evidence that people’s happiness forms a U shape over their lifetime, hitting its lowest point in midlife.
In the paper, Blanchflower reviewed large data sets from every continent but Antarctica. He drew from studies that focused on many measures of happiness, including life satisfaction, mood, pain, and more. He ran analyses that took into account factors that might impact happiness—such as health, income, employment, being married, being a parent, and more—then ran the analysis again not taking these into account.
In each case, he found that people’s happiness rose and fell in a U-shaped curve, and that it hit a low around the ages of 47 and 49. The consistency of this result surprised him.
“The expectation was that I probably wouldn’t find that this [dip in happiness] was the same everywhere; but it’s present in America, Germany, Thailand, Pakistan, everywhere,” he says. “Even in countries with a lower life expectancy, where you might’ve thought that it would be pretty different, it turns out it’s not. There’s an empirical regularity in the data from around the world, for millions of people.”
Why would that be? No one really knows, says Blanchflower; but researchers are trying to figure it out. The fact that apes also experience this drop in midlife points to a possible biological origin, perhaps with some evolutionary advantage. Feeling down in midlife could be nature’s way of spurring us to take action that extends our chances of survival or helps our communities.
«
Perhaps they’re disappointed by headlines that promise to answer a question and then say it doesn’t have an answer. Just throwing it out there.
unique link to this extract
Endangered migrating eagles impacted by Ukraine war • University of East Anglia
»
Researchers from the University of East Anglia (UEA), the British Trust for Ornithology (BTO) and the Estonian University of Life Sciences compared the movement and migration of the Greater Spotted Eagle through Ukraine, before and shortly after it was invaded by Russia in February 2022.
They were already studying the species when the war started, with the dangers faced by migratory birds usually related to disruptive weather or drought, changes in land use affecting traditional stopping-off points, or destruction of essential habitats.
However, during the invasion the team found that the eagles, which had previously been fitted with GPS tracking devices, were exposed to multiple conflict events along their journey through the country in March and April.
These included artillery fire, jets, tanks and other weaponry, as well as unprecedented numbers of soldiers moving through the landscape and millions of civilians displaced.
Their migratory behaviour, gathered from the tracking data, was compared to previous years as they passed between wintering areas in southern Europe and East Africa and key breeding grounds in southern Belarus.
Published on Tuesday in the journal Current Biology, the findings reveal the eagles made large deviations from their traditional migratory routes. They also spent less time stopping at their usual refuelling sites in Ukraine or avoided them entirely.
This resulted in the eagles travelling further and arriving on their nesting grounds later than usual. This could seriously affect them and likely contributed to reduced physical fitness at a time when peak condition is critical to successful breeding.
«
The animals think it’s stupid, just as the humans do, yet are equally powerless to stop it.
unique link to this extract
Microsoft Copilot will watch you play Minecraft, tell you what you’re doing wrong • Ars Technica
Kyle Orland:
»
Longtime gamers (and/or Game Grumps fans) likely know that even single-player games can be a lot more fun with a friend hanging out nearby to offer advice, shoot the breeze, or just offer earnest reactions to whatever’s happening on screen. Now, Microsoft is promising that its GhatGPT-4o-powered Copilot system will soon offer an imitation of that pro-social experience even for Minecraft players who don’t have any human friends available to watch them play.
In a pair of social media posts on Monday, Microsoft highlighted how “real-time conversations with your AI companion copilot” can enhance an otherwise solitary Minecraft experience. In the first demo, the disembodied copilot voice tells the player how to craft a sword, walking him through the process of gathering some wood or stone to go with the sticks sitting in his inventory. In another, the AI identifies a zombie in front of the player and gives the (seemingly obvious) advice to run away from the threat and “make sure it can’t reach you” by digging underground or building a tower of blocks.
These kinds of in-game pointers aren’t the most revolutionary use of conversational AI—even a basic in-game tutorial/reference system or online walkthrough could deliver the same basic information, after all. Still, the demonstration stands out for just how that information is delivered to the player through a natural language conversation that doesn’t require pausing the gameplay even briefly.
The key moment highlighting this difference is near the end of one of the video demos, when the Copilot AI offers a bit of encouragement to the player: “Whew, that was a close one. Great job finding shelter!” That’s the point when the system transitions from a fancy voice-controlled strategy guide to an ersatz version of the kind of spectator that might be sitting on your couch or watching your Twitch stream. It creates the real possibility of developing a parasocial relationship with the Copilot guide that is not really a risk when consulting a text file on GameFAQs, for instance (though I think the Copilot reactions will have to get a bit less inane to really feel like a valued partner-in-gaming).
«
Even so: the possibility of parasocial relationships with your automagically talking device does seem like something we should consider carefully.
unique link to this extract
Revealed: Meta approved political ads in India that incited violence • The Guardian
Hannah Ellis-Petersen:
»
The Facebook and Instagram owner Meta approved a series of AI-manipulated political adverts during India’s election that spread disinformation and incited religious violence, according to a report shared exclusively with the Guardian.
Facebook approved adverts containing known slurs towards Muslims in India, such as “let’s burn this vermin” and “Hindu blood is spilling, these invaders must be burned”, as well as Hindu supremacist language and disinformation about political leaders.
Another approved advert called for the execution of an opposition leader they falsely claimed wanted to “erase Hindus from India”, next to a picture of a Pakistan flag.
The adverts were created and submitted to Meta’s ad library – the database of all adverts on Facebook and Instagram – by India Civil Watch International (ICWI) and Ekō, a corporate accountability organisation, to test Meta’s mechanisms for detecting and blocking political content that could prove inflammatory or harmful during India’s six-week election.
According to the report, all of the adverts “were created based upon real hate speech and disinformation prevalent in India, underscoring the capacity of social media platforms to amplify existing harmful narratives”.
The adverts were submitted midway through voting, which began in April and would continue in phases until 1 June. The election will decide if the prime minister, Narendra Modi, and his Hindu nationalist Bharatiya Janata party (BJP) government will return to power for a third term.
«
As should always pointed out: dehumanising language (“vermin”) and othering is how massacres are fomented. Most recently in Myanmar, where this sort of monologue directed by Buddhists towards Muslims on, yes, Facebook, led to genocide in 2016-17.
unique link to this extract
AI traces mysterious metastatic cancers to their source • Nature
Smitri Mallapaty:
»
To treat metastatic cancers, doctors need to know where they came from. The origin of up to 5% of all tumours cannot be identified, and the prognosis for people whose primary cancer remains unknown is poor.
One method used to diagnose tricky metastatic cancers relies on tumour cells found in fluid extracted from the body. Clinicians examine images of the cells to work out which type of cancer cell they resemble. For example, breast cancer cells that migrate to the lungs still look like breast cancer cells.
Every year, of the 300,000 people with cancer who are newly treated at the hospital affiliated with Tianjin Medical University (TMU) in China, some 4,000 are diagnosed using such images, but around 300 people remain undiagnosed, says Tian Fei, a colorectal cancer surgeon at TMU.
Tian, Li Xiangchun, a bioinformatics researcher who studies deep learning at TMU, and their colleagues wanted to develop a deep-learning algorithm to analyse these images and predict the origin of the cancers. Their results were published in Nature Medicine on 16 April.
The researchers trained their AI model on some 30,000 images of cells found in abdominal or lung fluid from 21,000 people whose tumour of origin was known. They then tested their model on 27,000 images and found there was an 83% chance that it would accurately predict the source of the tumour. And there was a 99% chance that the source of the tumour was included in the model’s top three predictions.
…The researchers also retrospectively assessed a subset of 391 study participants some four years after they had had cancer treatment. They found that those who had received treatment for the type of cancer that the model predicted were more likely to have survived, and lived longer, than participants for whom the prediction did not match.
«
I feel this is a far better use for ML than helping people writing marketing letters.
| • 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 