Start Up No.1667: much more from the Facebook Papers, Iran suffers fuel cyberattack, a real giant iPod, Cobol’s problem, and more

Expect to see a lot of Tesla Model 3s on offer if you’re hiring a car from Hertz in the future – it just ordered 100,000 of them, more than a tenth of its existing fleet. CC-licensed photo by Raymond Wambsgans 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 12 links for you. Never cleaner. I’m @charlesarthur on Twitter. Observations and links welcome.

How Facebook failed the world • The Atlantic

Ellen Cushing:


Reading these documents [the Facebook Papers/Files] is a little like going to the eye doctor and seeing the world suddenly sharpen into focus. In the United States, Facebook has facilitated the spread of misinformation, hate speech, and political polarization. It has algorithmically surfaced false information about conspiracy theories and vaccines, and was instrumental in the ability of an extremist mob to attempt a violent coup at the Capitol. That much is now painfully familiar.

But these documents show that the Facebook we have in the United States is actually the platform at its best. It’s the version made by people who speak our language and understand our customs, who take our civic problems seriously because those problems are theirs too. It’s the version that exists on a free internet, under a relatively stable government, in a wealthy democracy. It’s also the version to which Facebook dedicates the most moderation resources. Elsewhere, the documents show, things are different. In the most vulnerable parts of the world—places with limited internet access, where smaller user numbers mean bad actors have undue influence—the trade-offs and mistakes that Facebook makes can have deadly consequences.

According to the documents, Facebook is aware that its products are being used to facilitate hate speech in the Middle East, violent cartels in Mexico, ethnic cleansing in Ethiopia, extremist anti-Muslim rhetoric in India, and sex trafficking in Dubai. It is also aware that its efforts to combat these things are insufficient. A March 2021 report notes, “We frequently observe highly coordinated, intentional activity … by problematic actors” that is “particularly prevalent—and problematic—in At-Risk Countries and Contexts”; the report later acknowledges, “Current mitigation strategies are not enough.”

In some cases, employees have successfully taken steps to address these problems, but in many others, the company response has been slow and incomplete. As recently as late 2020, an internal Facebook report found that only 6% of Arabic-language hate content on Instagram was detected by Facebook’s systems. Another report that circulated last winter found that, of material posted in Afghanistan that was classified as hate speech within a 30-day range, only 0.23% was taken down automatically by Facebook’s tools. In both instances, employees blamed company leadership for insufficient investment.


Insufficient investment by a company that posted $9bn in profits in its latest quarter. And Cushing’s point, that the version Americans get is the very best version – the most attuned to their customs, their mores, their ethos – is all the more terrifying when you mull it over.

Other versions of this story: Politico; NY Times on India; WSJ on India.

By the way, the Guardian is now included in the Facebook Papers sharing group. Apparently American publications didn’t like the idea of it being involved at the start; the exclusion wasn’t Haugen’s idea.
unique link to this extract

Facebook prioritized ‘angry’ emoji reaction posts in news feeds • The Washington Post

Jeremy Merrill and Will Oremus:


Five years ago, Facebook gave its users five new ways to react to a post in their news feed beyond the iconic “like” thumbs-up: “love,” “haha,” “wow,” “sad” and “angry.”

Behind the scenes, Facebook programmed the algorithm that decides what people see in their news feeds to use the reaction emoji as signals to push more emotional and provocative content — including content likely to make them angry. Starting in 2017, Facebook’s ranking algorithm treated emoji reactions as five times more valuable than “likes,” internal documents reveal. The theory was simple: Posts that prompted lots of reaction emoji tended to keep users more engaged, and keeping users engaged was the key to Facebook’s business.

Facebook’s own researchers were quick to suspect a critical flaw. Favoring “controversial” posts — including those that make users angry — could open “the door to more spam/abuse/clickbait inadvertently,” a staffer, whose name was redacted, wrote in one of the internal documents. A colleague responded, “It’s possible.”

The warning proved prescient. The company’s data scientists confirmed in 2019 that posts that sparked angry reaction emoji were disproportionately likely to include misinformation, toxicity and low-quality news.


Not widely known: Facebook introduced the change because of users in Myanmar, who only had the option to “Like”, which they dutifully did (they thought it meant the same as “I’ve read this”) on everything – especially hate speech. (As I document in Social Warming.)

Facebook also couldn’t read what they were writing in posts because the Zawgyi font they used isn’t Unicode and doesn’t translate. So it needed icons – emoji – to see what they were saying. And then things got worse.
unique link to this extract

Will the SEC add to Facebook’s woes? • The New York Times

Cecilia Kang:


Whistle-blowers have filed at least nine complaints to the agency, which has oversight of public companies like Facebook, using a selection of the internal documents to argue that Facebook misled investors with a rosier picture of the company than they knew to be true. The Securities and Exchange Commission (SEC) can impose big fines for misleading investors and impose restrictions on corporate leaders.

A case from securities regulators is probably far from a slam dunk, several legal experts said. The accusations in the complaints don’t appear to be quite as clear-cut as many other accounting and fraud cases taken up by the agency, they said.

The S.E.C. declined to say whether it had opened an investigation. Even such a move could be problematic for Facebook, leading to depositions of top executives and forcing the company to share private communications and other business documents.

But to win a lawsuit accusing the company of misleading investors, regulators would have to prove that executives had intended to hide or lie about problems. Regulators would also have to prove that the information revealed by Ms. Haugen, or turned up in an investigation, could have changed trading or voting decisions by shareholders if it had been shared.

It would be even more difficult to hold top executives personally responsible. Regulators would have to demonstrate that Mr. Zuckerberg or other executives had explicit knowledge that Facebook was hiding or lying about information that could sway investors.

“The argument that Facebook prioritized profits isn’t convincing, because that’s what companies do,” said Howard Fischer, a former trial lawyer for the S.E.C. “There will very likely be an investigation because it’s so high profile, but it’s hard to see a clear case.”


It’s unlikely unless they get someone on tape saying they’re planning to con investors, basically. The SEC would be cautious of overreach.
unique link to this extract

Facebook removed the News Feed algorithm in an experiment. Then it gave up • Big Technology

Alex Kantrowitz:


Everyone seems to hate Facebook’s News Feed algorithm. Critics want to do away with it, and Congress may strip Facebook’s legal protections for the content it amplifies. Yet only now, after years of speculation, do we have an idea of what Facebook would look like without it. 

In February 2018, a Facebook researcher all but shut off the News Feed ranking algorithm for .05% of Facebook users. “What happens if we delete ranked News Feed?” they asked in an internal report summing up the experiment. Their findings: without a News Feed algorithm, engagement on Facebook drops significantly, people hide 50% more posts, content from Facebook Groups rises to the top, and — surprisingly — Facebook makes even more money from users scrolling through the News Feed.

…Turning off the News Feed ranking algorithm, the researcher found, led to a worse experience almost across the board. People spent more time scrolling through the News Feed searching for interesting stuff, and saw more advertisements as they went (hence the revenue spike). They hid 50% more posts, indicating they weren’t thrilled with what they were seeing. They saw more Groups content, because Groups is one of the few places on Facebook that remains vibrant. And they saw double the amount of posts from public pages they don’t follow, often because friends commented on those pages. “We reduce the distribution of these posts massively as they seem to be a constant quality compliant,” the researcher said of the public pages.

The experiment was still in progress at the time of the report, but the researcher indicated that things were going poorly. “Things are trending down,” they said in an internal forum, adding a chart showing declining user sessions. Even though the researcher kept the “integrity pass” in place, or the first layer of the algorithm that sorts for integrity ahead of engagement, they said that “integrity bad metrics still shot through the roof.”


Counterintuitive. But worth knowing.
unique link to this extract

Given everything that’s going on with the Facebook papers, maybe now is the time to buy Social Warming, my latest book, about the toxic effects of social networks.

A prototype original iPod • Panic Blog

Cabel Sasser:


there was lots of portable MP3 playing competition (like the titular Nomad in “less space than a Nomad“), but the iPod was one of the first times Apple showed up and did what we now think of as their standard move — they made The Apple Version®. It was personal, well-designed, innovative, meaningful, the sum of which was more than specs and checklists. We (I? The industry?) needed that. I have fond memories of Dave (who now works on Playdate) reverse-engineering the iPod database storage format so that you could use Audion to load songs onto it. I remember how plain fun it was to use — that click wheel, the original fidget toy! It was cool that I could use it as a tiny portable hard drive. The iPod was really good.

To celebrate, I want to show you something you’ve never seen before.

Now, there are a lot of mysteries in the Panic Archives (it’s a closet) but by far one of the most mysterious is what you’re seeing for the first time today: an original early iPod prototype.

We don’t know much about where it came from. But we’ve been waiting 20 years to share it with you.

…Clearly, this revision of the prototype was very close to the internals of the finished iPod. In fact, the date there — September 3rd, 2001 — tells us this one was made barely two months before it was introduced.


It’s absolutely HUGE – about nine times the size of the actual first iPod, and at least twice the depth – and a real horrible banana yellow. Verified as real by Tony Fadell, who came to Apple with the iPod idea.
unique link to this extract

Why you should read “Read Max” by Max Read • Substack

The aforementioned Max Read:


A few weeks ago, Adam Mosseri, the head of Instagram, told Recode’s Peter Kafka “We know that more people die than would otherwise because of car accidents, but by and large, cars create way more value in the world than they destroyed. And I think social media is similar.” Instagram and its parent company, Facebook, had been the subjects of a series of scathing Wall Street Journal articles, the most striking of which demonstrated that — as the Journal’s headline put it — “Facebook Knows Instagram Is Toxic for Teen Girls,” and Mosseri was attempting to explain that while, yes, Instagram might “destroy value” (“value” understood here to be, I guess, the emotional well-being of young women?) ultimately, its cosmic balance-sheet is in the black, presumably because those young women will buy things from advertisers to make themselves feel better.

This was a classically Facebook self-defense, by which I mean it was (1) glib, (2) disingenuous, and (3) utterly unconvincing. For starters, you need a fucking license to drive a car, and Mosseri’s comparison leads pretty easily to the conclusion that, for Facebook’s value proposition to make any sense, anyone who wants to have a Facebook account should be required to take a timed written test, and then a quick spin around a mocked-up licensing version of the site with a government proctor watching over their shoulder, to ensure that they weren’t a danger to themselves or others. 

On the podcast, Kafka pointed out that the auto industry is heavily regulated, and Mosseri said that Facebook would welcome regulation, which, as a well-resourced incumbent, it no doubt would, at least on the provider side. (Poster licenses didn’t come up, as far as I could tell.) But even within their relatively onerous regulatory framework, cars have a staggering death toll: they kill more than a million people a year directly, and countless more through their contributions to carbon emissions and good old-fashioned particulate pollution. This is still not really kind of technology that I think I would seek to compare my product to, though I recognize that for Facebook “cars” represents an upgrade from “cigarettes” or “heroin.”


Now people are starting to get properly angry about this.
unique link to this extract

The climate denial is coming from inside Facebook’s house • Gizmodo

Brian Kahn:


In the midst of the hottest October in human history, a question popped up on an internal Facebook message board. “Policy for Misinformation – Climate Change Denial?”

The question sparked a discussion, including with an employee arguing that Facebook allowing climate denial posts to run unchecked on the platform made sense because the science around a specific type of ulcer once shifted.

The post, available here, is part of a tranche of documents released by whistleblower Francis Haugen’s legal team that Gizmodo and other outlets have received access to. (You can see what we’ve turned up so far.) The names of “low-level” Facebook employees are redacted, so it’s unclear who specifically engaged in the debate over climate change denial content. But the chats are illuminating in just how hands-off Facebook has been with climate denial, and how even within a company committed to net zero emissions by 2030, a laissez-faire attitude about perpetuating denial still reigns in some corners.

The internal logs are from 2019, a year before Facebook opened its climate science information center page for business.


As Gizmodo points out, climate change has real effects. But there are still people in Facebook – who posted in response to the question – prepared to say “welllll, scientific consensus has changed in the past, you know.” And then, of course, allow it even though it’s misinformation.
unique link to this extract

Iran says cyberattack causes widespread disruption at gas stations • Reuters


A cyberattack disrupted the sale of heavily subsidised gasoline in Iran on Tuesday, state media reported, causing long queues at gas stations across the country weeks before the anniversary of 2019 street protests that followed fuel price hikes.

Iran says it is on high alert for online assaults, which it has blamed in the past on its arch-foes United States and Israel. The United States and other Western powers meanwhile have accused Iran of trying to disrupt and break into their networks.

“The disruption at the refuelling system of gas stations … in the past few hours, was caused by a cyberattack,” state broadcaster IRIB said. “Technical experts are fixing the problem and soon the refuelling process…will return to normal.”

The oil ministry said only sales with smart cards used for cheaper rationed gasoline were disrupted and clients could still buy fuel at higher rates, the ministry’s news agency SHANA reported.

“This attack was probably carried out by a foreign country. It is too early to announce by which country and in which way it was done,” Abolhassan Firouzabadi, secretary of Iran’s Supreme Council of Cyberspace, told state TV.


Very unclear though whether this really is foreign hackers, or dissidents inside Iran. It would be a strange thing for a foreign state to do.
unique link to this extract

New York Times journalist Ben Hubbard hacked with NSO’s Pegasus after reporting on previous hacking attempts • The Citizen Lab

Bill Marczak, John Scott-Railton, Siena Anstis, Bahr Abdul Razzak, and Ron Deibert:


New York Times journalist Ben Hubbard was repeatedly targeted with NSO Group’s Pegasus spyware over a three-year period from June 2018 to June 2021. The targeting took place while he was reporting on Saudi Arabia, and writing a book about Saudi Crown Prince Mohammed bin Salman.

The targeting resulted in Pegasus infections in July 2020 and June 2021. Notably, these infections occurred after Hubbard complained to NSO Group that he was targeted by the Saudi-linked KINGDOM Pegasus operator in June 2018.

While we attribute the 2020 and 2021 infections to NSO Group’s Pegasus spyware with high confidence, we are not conclusively attributing this activity to a specific NSO Group customer at this time. However, we believe that the operator responsible for the 2021 hack is also responsible for the hacking of a Saudi activist in 2021.

Some forensic artifacts that we connect to NSO Group are present on Hubbard’s device as early as April 2018, although we are unable to confirm whether this represents a genuine infection attempt or a feasibility test.

A phone number belonging to Hubbard also reportedly appeared on the Pegasus Project list in July 2019. Unfortunately, forensic evidence is not available for this timeframe.


Unlike the Iran hack, it’s staringly obvious who was behind this one, and why. It’s remarkable how despite having no official diplomatic relations, there’s technology exchange (mostly one-way: from Israel to Saudi Arabia) against a common enemy, Iran.

And, of course, New York Times journalists.
unique link to this extract

Hertz orders 100,000 Tesla Model 3 Cars for rental fleet; Tesla stock surges • Bloomberg

Erik Schatzker:


Hertz Global Holdings Inc., barely four months out of bankruptcy, placed an order for 100,000 Teslas in the first step of an ambitious plan to electrify its rental-car fleet. Tesla’s shares soared, pushing the automaker’s value past $1 trillion for the first time.

The cars will be delivered over the next 14 months, and Tesla’s Model 3 sedans will be available to rent at Hertz locations in major U.S. markets and parts of Europe starting in early November, the rental company said in a statement. Customers will have access to Tesla’s network of superchargers, and Hertz is also building its own charging infrastructure.

It’s the single-largest purchase ever for electric vehicles, or EVs, and represents about $4.2bn of revenue for Tesla, according to people familiar with the matter who declined to be identified because the information is private. While car-rental companies typically demand big discounts from automakers, the size of the order implies that Hertz is paying close to list prices. 

“How do we democratize access to electric vehicles? That’s a very important part of our strategy,” Mark Fields, who joined Hertz as interim chief executive officer earlier this month, said in an interview. “Tesla is the only manufacturer that can produce EVs at scale.”


18 months ago it was considering selling big chunks of its rental fleet of 730,000 vehicles. Which also shows you how significant this purchase is: that’s a big slice of the whole fleet. That comment about Tesla being the only EV maker at scale is significant too. Everyone else is miles (or kilometres) behind.
unique link to this extract

COBOLing together UI benefits: how delays in fiscal stabilizers impact aggregate consumption • SSRN

Michael Navarrete is from the department of Economics at the University of Maryland:


The United States experienced an unprecedented increase in unemployment insurance (UI) claims starting in March 2020, mainly due to layoffs caused by COVID-19. State unemployment insurance systems were inadequately prepared to process these claims. Those states using an antiquated programming language, COBOL, to process UI claims experienced longer delays in benefit disbursement.

Using daily card consumption data from Affinity Solutions, I employ a two-way fixed effects estimator to measure the causal impact of COBOL-induced delays in UI benefits on aggregate consumption.

In this paper, I show that while aggregate consumption (as measured by credit and debit card purchases of consumption goods) fell precipitously in all states at the start of the pandemic and remained below pre-pandemic levels for months, this drop in consumption was significantly larger in states that run their UI benefit systems using COBOL.

I estimate that average consumption in COBOL-states fell 4.4 percentage points more than non-COBOL states between mid-March and mid-September of 2020. Using this estimate, I calculate that the failure to invest in updating UI benefit systems in COBOL states caused U.S. real GDP to fall by an extra $181bn (in 2012 dollars) during this time period.


If that’s the cost to the states of using Cobol, imagine what it’s costing banks.
unique link to this extract

Tesco opens first “just walk out” store to take on Amazon • Retail Gazette

Sahar Nazir:


Tesco has opened its first “just walk out” store, giving customers the chance to buy groceries without having to scan items or visit a till.

The supermarket’s GetGo store has opened in Holborn, central London, and follows a small trial of a similar store at Tesco head office in Welwyn Garden City.

Weight sensors in the shelves work with an AI system that can track an individual’s movement around the store and monitor the items they pick up via cameras, which follow each shopper.

The AI system works by building a unique skeleton outline of each person rather than using facial recognition.

Shoppers must download the app to use the store, where they check in by scanning a QR code generated on their phone.

Once inside, shoppers can pick up the items they want without scanning them. The bill is automatically charged to a shopper’s Tesco account when they leave. Cigarettes and alcohol are sold in a separate zone where a member of staff, about a dozen of whom are on duty on any day, is present to check that shoppers meet age restrictions.


So Tesco gets just as much information as Amazon – who, what, when, where (well, only one “where” presently). The idea of the “skeleton outline” isn’t one I’d heard before.
unique link to this extract

Errata, corrigenda and ai no corrida: An error in yesterday’s edition: Frances Haugen was impressed by Jeff – not Ben – Horwitz of the WSJ.

1 thought on “Start Up No.1667: much more from the Facebook Papers, Iran suffers fuel cyberattack, a real giant iPod, Cobol’s problem, and more

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

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