Start Up No.2201: Israel’s AI targeting machine, Twitter to charge new users to post, be an energy patriot!, Intercept struggles, and more


What effect might it have on your smartphone use if you changed its display to grayscale? CC-licensed photo by Jannis Andrija Schnitzer 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. WOB or BOW, though? I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.


‘Lavender’: the AI machine directing Israel’s bombing spree in Gaza • 972 Mag/Local Call

Yuval Abraham and Amjad Iraqi:

»

In 2021, a book titled “The Human-Machine Team: How to Create Synergy Between Human and Artificial Intelligence That Will Revolutionize Our World” was released in English under the pen name “Brigadier General Y.S.” In it, the author — a man who we confirmed to be the current commander of the elite Israeli intelligence unit 8200 — makes the case for designing a special machine that could rapidly process massive amounts of data to generate thousands of potential “targets” for military strikes in the heat of a war. Such technology, he writes, would resolve what he described as a “human bottleneck for both locating the new targets and decision-making to approve the targets.”

Such a machine, it turns out, actually exists. A new investigation by +972 Magazine and Local Call reveals that the Israeli army has developed an artificial intelligence-based program known as “Lavender,” unveiled here for the first time. According to six Israeli intelligence officers, who have all served in the army during the current war on the Gaza Strip and had first-hand involvement with the use of AI to generate targets for assassination, Lavender has played a central role in the unprecedented bombing of Palestinians, especially during the early stages of the war. In fact, according to the sources, its influence on the military’s operations was such that they essentially treated the outputs of the AI machine “as if it were a human decision.”

Formally, the Lavender system is designed to mark all suspected operatives in the military wings of Hamas and Palestinian Islamic Jihad (PIJ), including low-ranking ones, as potential bombing targets. The sources told +972 and Local Call that, during the first weeks of the war, the army almost completely relied on Lavender, which clocked as many as 37,000 Palestinians as suspected militants — and their homes — for possible air strikes.

«

But that, as we know had terrible consequences.
unique link to this extract


Fake footage of Iran’s attack on Israel is going viral • WIRED

Vittoria Elliott:

»

In the hours after Iran announced its drone and missile attack on Israel on April 13, fake and misleading posts went viral almost immediately on X. The Institute for Strategic Dialogue (ISD), a nonprofit think-tank, found a number of posts that claimed to reveal the strikes and their impact, but instead used AI-generated videos, photos, and repurposed footage from other conflicts that showed rockets launching into the night, explosions, and even President Joe Biden in military fatigues.

Just 34 of these misleading posts received over 37 million views, according to ISD. Many of the accounts posting the misinformation were also verified, meaning they have paid X $8 per month for the ‘blue tick’ and their content is amplified by the platform’s algorithm. ISD also found that several of the accounts claimed to be open source intelligence (OSINT) experts, which has, in recent years, become another way of giving legitimacy to their posts.

One X post claimed that “WW3 has officially started,” and included a video seeming to show rockets being shot into the night—except the video was actually from a YouTube video posted in 2021. Another post claimed to show the use of the Iron Dome, Israel’s missile defense system, during the attack, but the video was actually from October 2023. Both these posts garnered hundreds of thousands of views in the hours after the strike was announced, and both originated from verified accounts. Iranian media also shared a video of the wildfires in Chile earlier this year, claiming it showed the aftermath of the attacks. This, too, began to circulate on X.

«

The past year has been like an experiment in: what if we create perverse incentives to get people to post things that go viral, even or especially if they’re untrue, by paying them and giving them more prominence for those viral things?

Answer: nothing good.

unique link to this extract


Elon Musk plans to charge new X users to enable posting • TechCrunch

Ivan Mehta:

»

Elon Musk is planning to charge new X users a small fee to enable posting on the social network and to curb the bot problem.

In reply to an X account that posted about changes on X’s website, Musk said charging a small fee to new accounts was the “only way” to stop the “onslaught of bots.”

“Current AI (and troll farms) can pass ‘are you a bot’ with ease,” Musk said, referring to tools like CAPTCHA.

While replying to another user, Musk later added that new accounts would be able to post after three months of creation without paying a fee.

As is the case with a lot of announcements related to the social platform, there are no details at the moment about when this policy will be applicable and what fees new users might have to pay.

Last October, X started charging new unverified users $1 per year in New Zealand and the Philippines. New free users signing up for the platform from these regions could read the posts but couldn’t interact with them. To post content, like, repost, reply, bookmark and quote posts, they had to pay a fee. Musk might apply a fee similar to other regions.

«

Musk did allude to this ages ago, and has put it off for unknown reasons. The obvious one is that it won’t make any appreciable difference: there are so many old accounts which can be taken over, or have already been taken over, that the bot farms have plenty of opportunity left.
unique link to this extract


The “moronic inferno” and “the fidgets,” OR why my phone is now black & white • Forking Paths

Brian Klaas:

»

I confess: I have the fidgets and I want to escape, to borrow the wonderful term from Wyndham Lewis, from “the moronic inferno” that so much of social media and scroll culture has become. Whether Jonathan Haidt is right—or not—about the effect that smartphones are having on children (I suspect he is), I find his “Pascal’s Wager of Smartphone Usage” persuasive: if he’s right that it rots brains and causes despair and death, then he could be saving us all from the devastating effects of mental mush. However, if Haidt and his fellow Cassandras, are wrong, well, then you’ve just freed up some time in your day. No harm, no foul.

So, I’m trying something new: I’ve turned my smartphone screen grayscale. Silly as it sounds, it makes it wonderfully boring, reducing its seductive allure. I have also reincarnated an old phone, wiped it clean, and put nothing but a messaging app on it, so my friends and enemies can call or text me, but there’s nothing else. (If you want to go for the nuclear option, buy a dumb phone).6

And, for the pesky apps that suck you into the dreaded quadrant of ennui [when you’re in a passive and closed-minded state], I’ve installed an app called “one sec” that delays my access for a few seconds as it patronizingly tells me “It’s time to take a deep breath.” (When I first did this, I was horrified at how instinctive my finger movements were when I unlocked the phone. I truly am a dopamine ape!).

I have not succeeded. I still spend too much time in the wrong quadrants. But I am determined to avoid the dark fate that awaits us if we passively drift through life as closed-minded consumers of that most dystopian word: content.

«

unique link to this extract


YouTube cracking down on third-party apps that block ads • 9to5 Google

Abner Li:

»

Following the ad blocker crackdown, YouTube is explicitly going after third-party — often mobile — apps that let viewers skip advertising.

YouTube announced today that it is “strengthening our enforcement on third-party apps that violate YouTube’s Terms of Service, specifically ad-blocking apps.”

Users will see a “The following content is not available on this app” error message or experience “buffering issues” when they try to play content though those alternative clients.

We want to emphasize that our terms don’t allow third-party apps to turn off ads because that prevents the creator from being rewarded for viewership, and Ads on YouTube help support creators and let billions of people around the world use the streaming service.

YouTube Premium, which hit 100 million subscribers in February, is offered as the solution for those that “prefer an entirely ad-free experience.”

The Google video site says it only allows “third-party apps to use our API when they follow our API Services Terms of Service.” YouTube previously went after “YouTube Vanced” in 2022.

Going forward, it will crack down on clients that violate these policies: “…when we find an app that violates these terms, we will take appropriate action to protect our platform, creators, and viewers.”

«

Hardly unreasonable. YouTube is a gigantic revenue source for Google.
unique link to this extract


Are movies becoming more derivative? • Stephen Follows

Stephen works in films:

»

This may surprise some, but since 2000, just over half of all movies released have been original screenplays.

The most common source for adapted screenplays was real-life events, accounting for almost a fifth of movies made between 2000 and 2023. (Typically, in these cases, the filmmakers will have paid for the rights to a nonfiction book or two that covered those events, but we will classify that as ‘based on real-life events’ in this analysis.)

Other sources include fictional books/articles (8.9%), previous movies (11.8%), stage productions (including plays, musicals, and dance performances) (1.5%), and TV/Web shows (0.9%). In the chart below, ‘Other’ includes myths, legends, poems, songs, games, toys, and more.

How has this changed over the years? Forty years ago, about the same proportion of movies being made were original screenplays as they are today. That’s quite surprising – both because I assume that many people expected it to be lower in recent years, but also because little stays the same in the film industry over such a long period of time.

But when we look at a time series by year, we can see that it hadn’t plateaued. During the late 1990s and 2000s, original screenplays declined markedly and only rose again in the 2010s.

«

It is peculiar; I wondered if franchises aren’t a source of “original” screenplays which nonetheless make the whole cinema experience feel, well, derivative.
unique link to this extract


‘Energy patriots’: new analysis shows greenest homes can more than halve energy imports • BusinessGreen News

James Murray:

»

homes with heat pumps, insulation, and EVs use less than half the imported energy of households reliant on fossil fuels

Growing numbers of ‘energy patriots’ are helping to curb the UK’s reliance on fossil fuel energy imports by adopting new clean technologies.

That is the conclusion of a new analysis from the Energy and Climate Intelligence Unit (ECIU) think tank, which calculated that homes that using electric heat pumps, insulation, and electric vehicles use less than half of the imported energy than a household reliant on gas and petrol.

The study showed how a typical household using a gas boiler and petrol car will be dependent on imports for almost 70% of its energy, totalling around 17MWh a year. In contrast, a household that is insulated to Energy Performance Certificate grade C, uses a heat pump, and has an electric car will use 45% of the energy imports of a household with a gas boiler and a petrol car, at around 7.5MWh a year.

Households that also deploy solar panels can cut their use of fossil fuel imports further to just 6MWh, or around a third of the fossil fuel imports associated with a typical household.

«

“Energy patriots” is an amazing framing: it’s the sort of phrase you’d use to gin up all the people who jolt at “take back control”. I can almost see the fake advert with a square-jawed colonel starting at the camera and demanding “Are YOU an energy patriot?”
unique link to this extract


When do we stop finding new music? A statistical analysis • Stat Significant

Daniel Parris:

»

A New York Times analysis of Spotify data [in 2018] revealed that our most-played songs often stem from our teenage years, particularly between the ages of 13 and 16.

This finding has personal resonance, as I remember my cultural preferences being easily influenced during my pre-teen and early teenage years. For instance, I was twelve when Green Day released their landmark “American Idiot” album, a work that proved monumental in my relationship to music. Listening to the album’s titular track felt like a supreme act of rebellion (for a twelve-year-old suburbanite). I was entranced by this song’s iconoclastic spirit—could they actually say, “f**k America?”      

But “American Idiot” wasn’t a true act of revolution. In fact, the album was produced and promoted by a multinational conglomerate with the intent of packaging seemingly transgressive pop-punk acts for my exact demographic. How was I so thoroughly seduced by this song? And yet, to this day, my visceral reaction to “American Idiot” is still one of euphoria, despite my cynicism. I guess I have no choice but to love this song forever (thanks to pre-teen me). 

Indeed, YouGov survey data indicates a strong bias toward music from our teenage years, a phenomenon that is consistent across generations. Every cohort believes that music was “better back in my day.”  

Ultimately, cultural preferences are subject to generational relativism, heavily rooted in the media of our adolescence. It’s strange how much your 13-year-old self defines your lifelong artistic tastes. At this age, we’re unable to drive, vote, drink alcohol, or pay taxes, yet we’re old enough to cultivate enduring musical preferences. 

The pervasive nature of music paralysis across generations suggests that the phenomenon’s roots go beyond technology, likely stemming from developmental factors. So what changes as we age, and when does open-eardness decline?

Survey research from European streaming service Deezer indicates that music discovery peaks at 24, with survey respondents reporting increased variety in their music rotation during this time. However, after this age, our ability to keep up with music trends typically declines…

«

I wonder if it depends on the radio stations you listen to? Being exposed to different (new) musical styles makes a big difference.
unique link to this extract


Some ex-TikTok employees say the social media service worked closely with its China-based parent despite claims of independence • Fortune

Aledandra Sternlicht:

»

Evan Turner, who worked at TikTok as a senior data scientist from April to September in 2022, said TikTok concealed the involvement of its Chinese owner during his employment. When hired, Turner initially reported to a ByteDance executive in Beijing. But later that year, after the company announced a major initiative to store TikTok’s US user data only in the US, Turner was reassigned—on paper, at least—to an American manager in Seattle, he says. But Turner says a human resources representative revealed during a video conference call that he would, in reality, continue to work with the ByteDance executive. The stealth chain of command contradicted what TikTok’s executives had said about the company’s independence from ByteDance, Turner says.

Turner says he never met with the Seattle-based manager. Instead, Turner had weekly check-ins lasting less than seven minutes with the Beijing-based ByteDance executive. In these meetings, Turner says he merely told the executive how far along he was in completing assigned tasks—and nothing else.

Nearly every 14 days, as part of Turner’s job throughout 2022, he emailed spreadsheets filled with data for hundreds of thousands of US users to ByteDance workers in Beijing. That data included names, email addresses, IP addresses, and geographic and demographic information of TikTok U.S. users, he says. The goal was to sift through the information to mine for insights like the geographical regions where users watched the most videos of a particular genre and decide how the company should invest to encourage users to be more active. It all took place after the company had started its initiative to keep sensitive US user data in the US, and only available to US workers.

«

TikTok not doing what it said? Perish the thought.
unique link to this extract


The Intercept is running out of cash • Semafor

Max Tani:

»

The Intercept, the left-wing U.S. newsroom that’s been a thorn in Joe Biden’s side and a hub for pro-Palestinian coverage, is nearly out of money and facing its own bitter civil war, with multiple feuding factions battling for power and two star journalists trying to take control.

At the heart of the crisis is a nonprofit whose founding donor, Pierre Omidyar, decided in late 2022 to end his support for the organization. Now spun off from Omidyar’s First Look Media, The Intercept is losing roughly $300,000 a month, is on track to have a balance of less than a million dollars by November — and could be completely out of cash by May 2025, according to data shared internally in March.

The Intercept’s CEO, Annie Chabel, told Semafor in an interview this week that those projections were a worst-case scenario, and that the Intercept had a “stretch revenue goal that would allow us to continue into a longer horizon.”

The Intercept was born a decade ago in a very different moment for media and politics. Two of its founders, Glenn Greenwald and Laura Poitras, broke the story of Edward Snowden’s leaked surveillance files in 2013, which reshaped how Americans thought about the government and their privacy. Omidyar, a leftish billionaire with no known appetite for political combat, rapidly pledged deep support for an organization that would combine that anti-establishment mission with a combative form of online journalism born out of Gawker Media.

A decade later, American politics are almost unrecognizable. Greenwald quit in fury to make quixotic allies on the right. Liberal donors have lost their taste for party infighting as the spectre of Donald Trump looms, while voices further left are promising to punish Joe Biden over his response to Gaza.

«

It’s been relying on philanthropy, which doesn’t seem the smartest move in a media landscape that’s getting increasingly tight for funding.
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

Leave a comment

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