Algorithmic repetition: but how does Twitter’s function when fed a simalucrum of Trump’s tweets? CC-licensed photo by Ethan Miller 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 11 links for you. Use them wisely. I’m @charlesarthur on Twitter. Observations and links welcome.
Americans who mainly get their news on social media are less engaged, less knowledgeable • Pew Research Center
Amy Mitchell, Mark Jurkowitz, J. Baxter Oliphant and Elisa Shearer:
»
The rise of social media has changed the information landscape in myriad ways, including the manner in which many Americans keep up with current events. In fact, social media is now among the most common pathways where people – particularly young adults – get their political news.
A new Pew Research Center analysis of surveys conducted between October 2019 and June 2020 finds that those who rely most on social media for political news stand apart from other news consumers in a number of ways. These U.S. adults, for instance, tend to be less likely than other news consumers to closely follow major news stories, such as the coronavirus outbreak and the 2020 presidential election. And, perhaps tied to that, this group also tends to be less knowledgeable about these topics.
«
Also tend to be under 30, and have heard more conspiracy junk. Not investigated but likely: they’re more politically polarised than people who don’t use social media so much.
Would be great if there were comparative studies for other countries.
unique link to this extract
Relevant content: Twitter’s algorithm does not seem to silence conservatives • The Economist
»
The [US] president says that “social media platforms totally silence conservatives’ voices.” However, a study by The Economist finds the opposite. Twitter’s feed used to show people the latest posts from accounts they followed, but in 2016 it launched an algorithm to serve “relevant” tweets to users, even if they were days old and from unfamiliar accounts. We compared the two systems, and found that the recommendation engine appears to reward inflammatory language and outlandish claims.
Our experiment began in June 2019, when we created a clone of Mr Trump’s profile. This bot used his picture, biography and location, and followed the same people as he did. We used it to re-post some of the president’s old tweets over several weeks, so that the algorithm could learn what our Trump clone cared about.
Then from September to December we checked every ten minutes if Mr Trump had tweeted something. If so, three things happened. First, our clone repeated the tweet. Second, we checked its Twitter feed and recorded the first 24 posts served by the algorithm. Finally, we simulated what a chronological feed might have looked like, using the 24 most recent tweets by accounts that Mr Trump follows.
Our algorithmic and chronological feeds differed starkly. Nearly half the recommended tweets were from users whom Mr Trump does not follow. Using sentiment-analysis tools to extract feelings from text, we found the average curated tweet was more emotive, on every scale, than its chronological equivalent—and more so than Mr Trump’s own posts, too.
«
The algorithm is picking for engagement – and emotive words do that. There’s plenty of solid academic research on this. Suitable material for a book, really.
unique link to this extract
Twitter bans former KKK leader David Duke • The Washington Post
Jacob Bogage and Eugene Scott:
»
Avowed white supremacist David Duke was permanently banned from Twitter for repeated violations of the social media platform’s rules on hate speech.
The former Ku Klux Klan leader and one-time Louisiana legislator’s most recent tweets included a link to an interview he conducted with Holocaust denier Germar Rudolf. Other posts promised to expose the “systemic racism lie,” as well as the “incitement of violence against white people” by Jewish-owned media. He also shared misinformation about the danger and spread of the coronavirus.
“People who refuse the mask are the real heroes,” he tweeted.
Duke, who most recently is known for endorsing President Trump, was banned in June, which Twitter confirmed Thursday evening. He also was banned by YouTube that month.
Twitter and other social media platforms have been under fire for years for lax regulations on racist, anti-Semitic and misogynistic commentary from users, especially those who self-identify with hate groups.
«
Good to see Twitter focusing on the aim of reducing its user numbers to the nominal 250 million. Though seriously, this is long overdue. A platform that removed Graham Linehan (for his noise over the trans issue) before it removed David Duke needs to think about its priorities.
Obvious question: is Duke (still) on Facebook? Does he say the same things?
unique link to this extract
Three people have been charged for Twitter’s huge hack, and a Florida teen is in jail • The Verge
Sean Hollister:
»
Early on July 31st, the FBI, IRS, US Secret Service, and Florida law enforcement placed 17-year-old Graham Clark of Tampa, Florida, under arrest. He’s accused of being the “mastermind” behind the biggest security and privacy breach in Twitter’s history, one that took over the accounts of President Barack Obama, Democratic presidential candidate Joe Biden, Bill Gates, Elon Musk, Kanye West, Apple, and more to perpetrate a huge bitcoin scam on July 15th.
Apparently, he wasn’t alone: shortly after the Tampa arrest was revealed and after we published this story, two more individuals were formally charged by the US Department of Justice: 22-year-old Nima Fazeli in Orlando and 19-year-old Mason Sheppard in the UK. They go by the hacker aliases “Rolex” and “Chaewon,” respectively, according to the DOJ. The FBI says that two individuals in total are in custody. An unidentified minor in California also admitted to federal agents that they’d helped Chaewon sell access to Twitter accounts.
But according to an affidavit released late Friday, authorities have probable cause to believe Clark, the Tampa teen, was the one who got access to Twitter’s internal tools and directly carried out the scam. Specifically, he allegedly convinced a Twitter employee that he worked in the Twitter IT department and tricked that employee into giving him the credentials.
«
As others have observed, if this is what some script kiddies with plausible social engineering can do, what might state actors who are really working on it be up to? We know the Saudis infiltrated the company. What else might be happening?
unique link to this extract
AI-generated text is the scariest deepfake of all • WIRED
Renee DiResta:
»
undetectable textfakes—masked as regular chatter on Twitter, Facebook, Reddit, and the like—have the potential to be far more subtle, far more prevalent, and far more sinister. The ability to manufacture a majority opinion, or create a fake-commenter arms race—with minimal potential for detection—would enable sophisticated, extensive influence campaigns. Pervasive generated text has the potential to warp our social communication ecosystem: algorithmically generated content receives algorithmically generated responses, which feeds into algorithmically mediated curation systems that surface information based on engagement.
Our trust in each other is fragmenting, and polarization is increasingly prevalent. As synthetic media of all types—text, video, photo, and audio—increases in prevalence, and as detection becomes more of a challenge, we will find it increasingly difficult to trust the content that we see. It may not be so simple to adapt, as we did to Photoshop, by using social pressure to moderate the extent of these tools’ use and accepting that the media surrounding us is not quite as it seems. This time around, we’ll also have to learn to be much more critical consumers of online content, evaluating the substance on its merits rather than its prevalence.
«
I hardly ever disagree with DiResta, but I just don’t see why, this many years into social media and our knowledge of bots, anyone takes “comments on Facebook/Twitter” as a metric of anything. It’s so easy to fake – even before GPT-3 – that there’s no value in trusting it.
unique link to this extract
One flight is worth a thousand Big Macs: digesting these hard facts killed my appetite for flying • The Correspondent
Jelmer Mommers:
»
We really do have to acknowledge the hard truth. If we return to flying as we did before Covid-19, we’ll never bring global warming under control.
The urgent need to fly less may feel painful. It spells the end for a certain way of living. Since the ending of the second world war, flying has been associated with shifting boundaries, meeting people, new experiences and the discovery of new cultures. “Travel is fatal to prejudice, bigotry, and narrow-mindedness,” wrote Mark Twain. He was right.
In flight over planet earth, holidaymakers, business people, academics would all sit together, paging through the same magazines, listening to announcements from the cockpit. It was sweet.
But looking at the facts about flying and climate, we can arrive at only one conclusion. To keep the planet liveable, those of us who fly need to do it much less, or stop altogether.
Flying is a small but growing source of emissions. At just over 2% of global CO₂ emissions, it currently represents only a limited share of the total. But prior to coronavirus, this expanding industry was forecast to account for a fifth of all emissions by 2050.
…Perhaps you have already decided to give up meat for the sake of the climate: did you know that one return flight from London to New York is as bad for the climate as consuming almost 1,000 Big Macs? Have you swapped old lightbulbs at home for environmentally friendly LEDs? The CO₂ you will save over five years is cancelled out by one medium-haul flight, from, say, Berlin to Lisbon.
I’m not saying these things to suggest that vegetarianism and energy saving are pointless. On the contrary, these are effective steps to lower your carbon footprint. But here’s that uncomfortable truth again: if we continue to fly, we will undo the progress made in many other areas.
«
(Via John Naughton)
unique link to this extract
Why America is afraid of TikTok • The Atlantic
Michael Schuman:
»
Companies that operate in China, both local and foreign, repeatedly get into [political] scuffles with thin-skinned officials over perceived political incorrectness. But this type of interference has heightened scrutiny of TikTok’s decisions about content. All social-media outfits have challenges with content moderation, but with TikTok, critics make the assumption that choices about what should and should not be on the app are made to please Chinese censors. TikTok, of course, denies this, and insists that decisions are made by the U.S. team.
The inevitable controversies have ensued anyway. Critics accused TikTok of scrubbing videos supportive of prodemocracy protesters in Hong Kong, whom Beijing considers “terrorists,” and of locking the account of a teen who shared a video critical of the Chinese government’s ill-treatment of the country’s minority Uighur community. (The company has denied both accusations. In the former case, an investigation by BuzzFeed News backed up TikTok’s assertion, and in the latter one, TikTok said the video was not the reason the account was frozen, and it apologized and reinstated the user’s access.) In the U.S., following the death of George Floyd, TikTok came under fire for allegedly suppressing videos related to Black Lives Matter and the protests against police brutality and racism. The company has said that this was a temporary technical glitch.
Facing perceived threats from China, Congress and the White House have tended to confront them with measures American officials previously preferred to avoid—restrictions on businesses and people.
«
US to widen action against Chinese tech groups beyond TikTok • Financial Times
Aime Williams and Hannah Murphy:
»
The Trump administration has vowed to “take action” in a matter of days against Chinese software companies that it perceives as a risk to security, in a sign that Washington is set to broaden its offensive beyond the video-sharing app TikTok.
ByteDance, the Chinese owner of TikTok, is racing to save the app’s US operations with a plea to the administration to allow it to sell the unit to Microsoft.
Comments from US secretary of state Mike Pompeo on Sunday suggested that additional action against a wider range of Chinese technology companies would follow.
“These Chinese software companies doing business in the United States, whether it’s TikTok or WeChat — there are countless more . . . are feeding data directly to the Chinese Communist party, their national security apparatus,” Mr Pompeo told Fox News.
“President Trump has said ‘enough’ and we’re going to fix it and so he will take action in the coming days with respect to a broad array of national security risks that are presented by software connected to the Chinese Communist party.”
«
Pompeo is an pompous enabling windbag, but questioning what data gets collected is a start. Maybe someone will question the data that gets collected about all Americans and sold to data brokers, who can then sell it to intermediaries run by Chinese companies who could be beholden to the CCP in just the same way. If there’s a leak, water will flow.
unique link to this extract
Apple leaks reveal upcoming product launch dates • Seeking Alpha
SA Editor Yoel Minkoff:
»
It’s going to be a busy fall season for Apple, according to well-known leakers iHacktu Pro and Komiya, who published the launch dates for every upcoming company product.
The late 2020 updates will begin on August 19 with a new iMac, AirPods Studio, HomePod 2 and HomePod Mini, followed by an event on September 8 that will unveil the iPhone 12 line, iPad, Apple Watch Series 6 and AirTags.
Another special event on October 27 will show off the Apple Silicon MacBook and MacBook Pro 13″, iPad Pro and Apple TV 4K.
There’s also big expectations for a renewed AirPower charging mat, and smaller wireless charger AirPower Mini, as well as Apple Glass – the reported augmented reality smart glasses.
«
Just had to go over the edge with the “Apple Glass” thing at the end. (No. Way.) The dates and products otherwise sound… possible?, although the 19th is a Wednesday – Apple prefers Tuesdays. The August thing sounds a bit ambitious, but the October date for the MacBook (back with a bang, sorry, ARM, sorry, Apple Silicon chip) and similar MacBook Pro 13in all makes sense. Anyway, you could put the dates in your diary and see how they fare.
Related: Apple confirmed last week that the iPhone won’t arrive in September (though everyone expects it to be announced in September). Last times that’s happened was the iPhone XR in 2018, the iPhone X in 2017 (didn’t arrive until November), and before that the iPhone 4S in 2011 (launched October).
unique link to this extract
Apple vs Google: a tale of two ecosystems • Android Authority
Chruv Bhutani isn’t thrilled by the lack of integration between Android, Google, Chromebooks and WearOS, especially compared to Apple’s cross-device integration:
»
It could be argued that by licensing out its software and operating system Google is just an enabler for a broader ecosystem running on its platform, and that’s fair enough but you don’t invest a fortune, buy two smartphone companies, and a wearable manufacturer without having serious hardware ambitions. Between the Chrome OS running Pixelbooks, the Pixel series of phones, and Nest hardware, Google has been trying to create a semblance of an alternative to Apple’s hegemony, and with it comes the responsibility to do it right.
However, operating in silos with each product acting as a distinct vertical just hasn’t worked to Google’s advantage. This lack of a unified focus and unwillingness to listen to what the market demands was epitomized by the launch of the flawed Pixel 4 and the subsequent departure of key executives. This is all the more astonishing in a time when even the notoriously stubborn Apple is willing to budge and add widgets to iPhones and iPad.
Look, I get that Google can’t or doesn’t want to miff its partner relations. That doesn’t mean gimping your own hardware is acceptable, however.
«
Alphabet grows up – but Google’s problems are bigger than just the antitrust case • The Economist
»
These days employees are being told to access sensitive documents only if they “need to know”. Some staff talk of creating if not a labour union, then at least a group to defend their interests.
In the wake of the killing of George Floyd many Googlers criticised their top management for doing too little, too late to make the company more diverse; after a couple of weeks the firm vowed to raise the “leadership representation of underrepresented groups” by 30% over the next five years. In June more than 2,000 employees signed an open letter to Mr Pichai demanding that the company stop selling its technology to police forces across America.
Over the past few weeks things have seemed to calm down internally. But the respite may be superficial. Many workers are keeping their mouths shut for fear of being laid off, one Googler reports. Few relish the thought of losing a cushy job in a recession. Activists now shun the firm’s communication tools and organise elsewhere online.
All this fuels murmurings and speculation, both inside and outside Alphabet, over whether Mr Pichai is the right person for the job. Some Google executives and engineers describe him as “too checked out” and his leadership as “uninspired”. He is also accused of excessive risk aversion. “I’ve never shied away from making big bets and following my instincts,” Mr Pichai insists. But it is hard to argue that he has shown the vision of Amazon’s Jeff Bezos or Microsoft’s Satya Nadella.
«
He’s hardly in the position of Bezos, building from the ground up, or Nadella, trying to fix a very broken company. All he can easily do is screw things up; getting it right is harder.
unique link to this extract
Errata, corrigenda and ai no corrida: none notified
The Overspill: when there's more that I want to say 