Start Up No.1772: Musk gets Twitter board seat, Musk v the SEC, your lost photo masterpieces, Yandex’s Russian data questions, and more

Moon-faced sunblocker Steven Seagal has appeared in a number of junky films known in the trade as “geezer teasers”, which offer easy work for past-it action stars. CC-licensed photo by Gage Skidmore 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. Unboarded. I’m @charlesarthur on Twitter. Observations and links welcome.

Musk joins Twitter board in deal that prevents him from buying majority stake • Ars Technica

Jon Brodkin:


Elon Musk is joining Twitter’s board of directors in a deal that prohibits him from buying more than 14.9% of the company’s stock, Twitter announced today. The news of Musk joining the board comes one day after the Tesla and SpaceX CEO revealed that he had purchased 9.2% of Twitter shares.

Musk said he’s looking forward to helping Twitter make “significant improvements” and asked Twitter users in a poll if they want an edit button. After about 3.4 million votes, 73.4% of respondents had voted yes—or rather, they voted “yse” instead of “on” because Musk misspelled both options. [Gee, wonder if that was a jkoe – Overspill Ed.]

Twitter CEO Parag Agrawal suggested that Musk’s poll on an edit button might influence Twitter policy. “The consequences of this poll will be important. Please vote carefully,” Agrawal wrote in a retweet of the poll.

Though a potential downside of an edit button is that users could significantly change tweets that are going viral, Meta CTO Andrew Bosworth wrote, “We solved this on Facebook a long time ago. You just include an indicator that it has been edited along with a change log. If you are really worried about embeds, they can point to a specific revision in that history but with a link to the latest edit. Not a real issue.” Musk’s response to Bosworth said only that “Facebook gives me the willies.”

Musk’s agreement to join the board prevents him from taking a controlling stake in Twitter. “For so long as Mr. Musk is serving on the Board and for 90 days thereafter, Mr. Musk will not, either alone or as a member of a group, become the beneficial owner of more than 14.9% of the Company’s common stock outstanding at such time, including for these purposes economic exposure through derivative securities, swaps, or hedging transactions,” Twitter wrote in a Securities and Exchange Commission filing submitted today.


That’s if you think Musk sets any store by SEC filings – see the next link. Clearly the board seat is the quid pro quo for not buying a majority stake; it’s also a smart way for Musk of not getting thrown off Twitter. (“Imagine if Trump was actually a billionaire,” someone commented.) But really, it’s not good news. What expertise is he going to bring? They’re not making cars, solar panels or rockets. Will it be his knowledge of shitposting?
unique link to this extract

9.2% and the Master of Twitter • Margins

Ranjan Roy:


I’m sure most of you have followed Elon [Musk] vs. the SEC in some capacity. But I think the timeline context is important to understand where we are today.

The most important question for me – Why did Musk start escalating in February?

The SEC had treated the 2018 settlement with kid gloves so why force their hand? Yes, there was a new insider trading investigation and some subpoenas, but why push this fight? Since the 2018 settlement, amidst that status quo detente, TSLA is up nearly 1750% and Tesla, the business, feels like it’s at its most stable operating performance ever. Why choose this moment to mess with the very circumstances that drove some of the greatest wealth creation in human history?

I had the chance to speak about this entire drama on March 10th on TechCheck – around the 3:05 mark in the video I laid three potential ways this could play out.

• The SEC could fight tooth and nail just to maintain the status quo where they were not really regulating him. This seemed unlikely.

• The SEC could agree the original settlement is void and then go after him over every single tweet since (even the original 420 tweet). This seemed most likely.

• The most extreme possibility I saw was the SEC could try to get Elon Musk kicked off of Twitter.

I know, it sounds ridiculous that the SEC could try to stop Musk from tweeting at all.

But if someone uses their personal Twitter account to repeatedly break securities laws, it doesn’t seem too crazy that the SEC could force that person to stop tweeting. If someone specifically created a legally binding arrangement with the SEC, and then repeatedly violated that arrangement in order to break securities laws, it doesn’t seem like a 1st amendment issue if they lose their privilege to tweet.


He has plenty more. Roy is always worth reading. Effectively Musk is shortcircuiting the SEC’s power.
unique link to this extract

Jeff Bezos and Amazon just hired everybody but SpaceX for Project Kuiper • Ars Technica

Eric Berger:


Amazon on Tuesday announced the largest commercial launch deal ever. The company said it has finalized agreements with three different rocket companies for a total of 83 launches. The rockets will deploy a majority of Amazon’s low-Earth-orbit constellation of broadband satellites.

With this deal, Amazon has acquired an extraordinary amount of medium- and heavy-lift launch capacity over the next five years, procuring launches from every major Western provider except for its direct satellite competitor, SpaceX. Aside from SpaceX, this purchase represents the vast majority of any “spare” launch capacity for larger rockets in the United States or Europe over the next half-decade.

Amazon announced launch agreements with the following companies as it seeks to build out its constellation of 3,236 satellites:
• Arianespace: 18 launches of Europe’s new Ariane 6 rocket
• Blue Origin: 12 launches of the company’s New Glenn rocket, with options for 15 additional launches
• United Launch Alliance: 38 launches of the company’s Vulcan rocket

Additionally, Amazon previously announced that it has purchased the final nine Atlas V rocket launches from United Launch Alliance before that vehicle, which is powered by Russian engines, is retired.


So that’ll be Amazon, and Starlink, and there’s the Inmarsat one. Is there really room for three commercial satellite broadband providers? If there’s consolidation, then as many as two in three of those satellites will be surplus. Let’s hope they can be deorbited cleanly.
unique link to this extract

Your camera roll contains a masterpiece • The New Yorker

Michael Johnston:


The late Erich Hartmann, a past president of Magnum, once showed me his friend Henri Cartier-Bresson’s negatives and contact sheets, stored at the famous photo agency’s New York offices in rows of three-ring binders lined up on shelves. Sheet after sheet contained not a single photograph I recognized. Some worked, most didn’t—not even for H.C.B.

Happily, there’s another side to the equation. If you take enough photographs, it’s almost inevitable that you’ll eventually get an extraordinary one, for reasons you might not understand. Cartier-Bresson was a hunter in his youth, and photographers have often described his brand of street photography as a kind of “hunting,” but it might be more accurate to say that it was like fishing—a sport in which you can do a lot to optimize your chances but still can’t know for sure what you’re going to get. Chance is pretty much always in play. Sometimes everything comes together before the lens, and the visual world sorts itself within the frame, and you get a little gift. None of us really knows for sure if or when the magic’s going to happen.

Today, of course, we’re in the age of digital photography. Back in the eighties, I remember reading that six billion photographs were taken each year, a number that seemed as big as the ocean; currently, although exact numbers can’t be known, the world probably collects that many images every three and a half days. There’s a new way in which we can miss out on great photographs: they can be buried forever in the digital tsunami.


Statistically, of course this must be true. He advises that you should “scroll your roll”. I think we should leave it to AI to have a stab at finding the very best ones. That happens already on the iPhone to some extent – it produces “memories” – but I think it offers far too many. And of course it can’t know about emotional impact. Maybe it could be trained on the few photos we pick as “favourites”?
unique link to this extract

Russian–Ukrainian war • OpenStreetMap Wiki


On behalf of the OSM Ukraine in connection with Russia’s ongoing military invasion of Ukraine we are turning to people who are contributing to the development and improvement of the OSM data.

We urge everyone to refrain from any mapping of the territory of Ukraine at the moment!

The Russo-Ukrainian War is unfolding on many fronts including the information one. The possible use of open data by Russian invaders to plan attacks on military and civilian objects is one of the most important reasons why we ask you not to perform any mapping of objects in Ukraine.

We shall take action to amend (delete, modify, revert to the previous state etc.) any found cases of mapping related to military or critical social infrastructure facilities as well as contact the DWG and other OSMF working groups to ban the users who systematically make similar changes (more than one).

Such a request (demand) echoes the provisions of the Article 114-2 of the Criminal Code of Ukraine according to which “Dissemination of information on redeployment, movement or location of the Armed Forces of Ukraine or other military formations established in accordance with the laws of Ukraine, if it is possible to identify them on the ground, if such information is not published by the General Staff of the Armed Forces of Ukraine, committed under martial law or state of emergency, shall be punishable by imprisonment for a term of five to eight years”


unique link to this extract

Fast, the one-click payments service, is shutting down • Protocol

Veronica Irwin and Benjamin Pimentel:


Online payments service Fast announced Tuesday that it is closing its doors, a sudden, stunning end to a seemingly fast-growing ecommerce venture once considered a pandemic darling.

The one-click-checkout software maker will discontinue service of its Fast Checkout on Friday, CEO and co-founder Domm Holland said in a statement. “Sometimes trailblazers don’t make it all the way to the mountaintop,” he said.

Fast ran out of funds after failing to secure additional investment fast enough in what had become a tough fundraising environment, a Fast employee told Protocol.

“We waited too long and we ran out of money,” said the employee, who asked not to be identified because of the sensitivity of the situation. Fast “misjudged significantly” the mood in the VC community, he added: “What was acceptable revenue and burn and prospects for growth in the summer of 2021 looks looks very different in April of 2022.”

…The pandemic greatly accelerated innovation in online shopping, and several other companies created their own one-click-checkout systems, including Shopify and Bold Commerce. PayPal was always considered a direct competitor to Fast, while Amazon invented one-click online checkout so long ago that its patent has expired. Apple auto-fills payment information on Macs and iPhones, as does Google’s Chrome browser.


Yeah, really wouldn’t have thought this was a market where – Amazon’s patent having expired – there’s much space for making one-click checkout products. A week ago it claimed it was about to raise $100m. Guess there wasn’t a one-click for that.
unique link to this extract

How Randall Emmett found success in Hollywood • Vulture

Joshua Hunt:


I first heard of [multiple producer and two-times director] Randall Emmett last September, while speaking with Adam Champ, an executive at Daro Film Distribution in Monaco. From his office in Côte d’Azur’s sun-drenched tax haven, Champ explained an inglorious but profitable slice of the film industry that is built around a certain category of actor — the kind of action stars and leading men who once ruled Hollywood and now make very good money appearing in very bad movies, most of them relegated to streaming services, video on demand (VOD), and late-night television in Europe and South America.

Among these actors are John Travolta, Nicolas Cage, and Sylvester Stallone. But perched atop the ignominious heap is Bruce Willis, whose prolific partnership with EFO Films, one of the biggest players in this niche of the industry, results in as many as four or five movies each year.

“With Bruce Willis, there’s almost a model for how he features in these movies,” Champ theorized. “One of my clients calls it a ‘geezer teaser’: You have Bruce Willis at the intro of the movie, so people are like, Great, this is a Bruce Willis movie. But he’s actually a secondary character who shows up sporadically.”

In most of Willis’s movies for EFO, “sporadic” would be a generous appraisal of his presence. The actor clocks just seven minutes of screen time in Hard Kill, and in Extraction, he spends less than nine minutes onscreen. In the home-invasion thriller Survive the Night, audiences get almost ten minutes out of the actor, even if they aren’t his best.


Willis’s brief appearances, of course, are now explicable as due to his aphasia. Travolta, Cage, Stallone and others such as Steven Seagal will have to find their own reasons. But the “geezer teaser” is a brilliant description for these films, which in a previous life would have been straight-to-DVD; now they’re straight-to-streaming. (Via Benedict Evans’s newsletter.)

Also perfect for this paragraph:


“If you throw Randy out the door, he comes in the window,” Lerner told the Los Angeles Times in 2012. “If you throw him from the window, he comes down the chimney.”


unique link to this extract

Russian tech giant Yandex’s data harvesting raises security concerns • Financial Times

Patrick McGee:


Yandex has acknowledged its software collects “device, network and IP address” information that is stored “both in Finland and in Russia”, but it called this data “non-personalised and very limited”. It added: “Although theoretically possible, in practice it is extremely hard to identify users based solely on such information collected. Yandex definitely cannot do this.”

The revelations come at a critical time for Yandex, often referred to as “Russia’s Google”, which has long attempted to chart an independent path without falling foul of Russian president Vladimir Putin’s desire for greater control of the internet.

The company said it followed “a very strict” internal process when dealing with governments: “Any requests that fail to comply with all relevant procedural and legal requirements are turned down.”

But Cher Scarlett, formerly a principal software engineer in global security at Apple, said once user information was collected on Russian servers, Yandex could be obliged to submit it to the government under local laws. Other experts said that the metadata of the sort collected by Yandex could be used to identify users.

…Yandex has software in the form of a software development kit, or SDK, called “AppMetrica”. SDKs are building blocks used by developers to create apps. The Google Maps SDK, for instance, allows apps to embed mapping functions rather than build that functionality from scratch. Many SDKs are offered for “free” in exchange for access to user data that aids targeted advertising.

Among the apps with AppMetrica installed are games, messaging apps, location-sharing tools and hundreds of virtual private networks — tools designed to allow people to browse the web without being tracked. Seven of the VPNs are made specifically for a Ukrainian audience. Total installs of apps that include the AppMetrica SDK are in the hundreds of millions, according to Appfigures, an app intelligence group.


Embedded SDKs really are both a necessity and a blight.
unique link to this extract

Stop saying Ukraine is winning the information war • The Atlantic

Carl Miller:


we compiled a roster of randomly selected accounts from across our new map [of tweets saying #IstandwithPutin and #IstandwithRussia] and delved into them, to try to draw out what set each of the different clusters of accounts apart. What struck us immediately was how clearly each cluster seemed to relate to geography—to the purported national identities and languages that the accounts used.

There was a dense knot of accounts identified as Indian that largely retweeted a stream of messaging in English and Hindi supporting Prime Minister Narendra Modi and his Hindu nationalist Bharatiya Janata Party. Another group used Urdu, Sindhi, and Farsi, with users primarily identifying as Iranian or Pakistani. One node was ostensibly from South Africa but included Ghanaian, Nigerian, and Kenyan users talking about public health, fuel shortages in Nigeria, and former South African President Jacob Zuma. A final cluster was the only one not characterized by language or geography. Accounts in this grouping sent the fewest tweets and had the fewest followers; many had been created either on the day of Russia’s invasion or on March 2, the day of a key United Nations vote condemning the invasion—and when I saw those hashtags suddenly trend.

…the early data are revealing, the activity suspicious. These accounts came alive for UN votes on the invasion, propelled in part, I suspect, by one or more “paid to engage” networks—groups of accounts that will shift their Twitter usage en masse to deliver retweets for a fee. But real people (we are unsure precisely how many) are also helping the hashtags trend. That interplay between organic and inauthentic activity is the most important subtlety of this research. It also gives us our most important conclusion.

Insofar as this was a coordinated campaign, we saw little attempt to address (or impersonate) Western social-media users. To the extent that we saw real people using the hashtag, very few were from the West.


unique link to this extract

I Was a Facebook content moderator. I am now living in a horror movie • Business Insider

Daniel Moutang:


I come from a large family in South Africa.

Shortly after graduating, around March 2019, I came across a content-moderation position at a company called Samasource (now Sama). The company and the profession was unknown to me at the time, but the company claimed its focus was on training poor people and lifting them out of poverty — so I applied.

I was quickly on the journey of a lifetime to work as a Facebook content moderator in Nairobi, Kenya.

It was an adventure that would change my life forever — especially because in my family, and village by extension, I was breaking records. I was the first one to go to a so-called prestigious university, travel in a plane and work abroad.

On that flight to Nairobi, I had no idea I would be working on social media — let alone on Facebook. Neither did I know that, while breaking those records, I would actually destroy my mental stability and physical health.

The job of content moderators is to try to make Facebook safe for everyone who uses it. Sadly, some of the billions of users on Facebook post horrible things every day — and our job is to sift through these posts and take down ones that violate Facebook’s rules so ordinary people don’t have to see them.

It’s gruelling work. Imagine long shifts in an office looking at a constant stream of videos and images of graphic violence, animals being tortured, and the sexual exploitation of children.

The first video I remember seeing was a livestream of someone being beheaded. I believe my mental health began to fray from that first video, and over time it got worse.


“We just want to connect everyone in the world.” But this becomes incredibly problematic. You have to have rules on a social media site. Yet equally, sometimes those rules flip too far: the atrocities in Bucha and other Ukrainian cities were being automatically flagged and their hashtags blocked on Facebook and Instagram because there was “graphic content” associated with it. AIs don’t know there’s a war on. They don’t know which side has the moral merit. They don’t know that we sometimes need to be shown the graphic content – or at least have the option of looking away.
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 Reply

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

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

Twitter picture

You are commenting using your Twitter 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.