Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Will Oremus Profile picture
Will Oremus
@WillOremus
writing about tech and its discontents for the @washingtonpost. focused on AI, algorithms, social media, and platforms. dm for signal.
Potato Of Reason Profile picture Yomi Shishio Profile picture 2 subscribed
Apr 19, 2023 • 9 tweets • 5 min read
This visual deep dive into one of the largest AI language datasets is nonstop fascinating and troubling and anyone who is remotely interested in how LLMs really work, their biases, or intellectual property should read it. washingtonpost.com/technology/int… "Content without consent" is a concern that I could see catching on as more people gradually realize the content they've published and posted over the years is being secretly used to train for-profit AI models. washingtonpost.com/technology/int… Image
Apr 5, 2023 • 7 tweets • 5 min read
Asked for examples of sexual harassment at law schools, ChatGPT named a GW prof accused of touching a student on a class trip to Alaska, citing a WashPost story.

The prof is real. The rest was made up.

We wrote about what happens when AIs lie about you: washingtonpost.com/technology/202… It gets weirder. Bear with me.

ChatGPT generated the fake scandal involving law prof @JonathanTurley in response to prompts from @VolokhC last week. Turley wrote about it in a @USATODAY op-ed Monday.

Today we tested the same prompt on Microsoft's Bing AI. And guess what...
Dec 16, 2022 • 5 tweets • 2 min read
remember when musk claimed he was a “free speech absolutist” Amazing. Now Twitter has suspended my colleague, Washington Post tech reporter Drew Harwell, without explanation--apparently for his tweet critical of Musk, which included a screenshot of Mastodon's tweet.
Dec 14, 2022 • 4 tweets • 2 min read
bizarre as it sounds, i think this is actually the clearest and most plausible explanation yet for how musk justifies this entire twitter boondoggle to himself musk saw twitter as the root of the “woke mind virus” which he blamed for everything from the media criticizing him to his employees complaining about racism & sexism to liz warren pushing a billionaire tax. and he thought buying it could put a stop to all that.
Dec 9, 2022 • 15 tweets • 6 min read
There are some mildly interesting details in here but presenting it as some new and shocking and secret finding is either ignorant or disingenuous. There have been whole news cycles about Twitter shadowbans. I know because I’ve written about them multiple times over the years. Sorry, I do have more to say about this, just have to finish putting the kid to bed first lol
Dec 8, 2022 • 6 tweets • 3 min read
New: The war has sent *8 million* Ukrainians fleeing to Europe. With Russia shelling infrastructure ahead of a frigid winter, more are on the way. Germans and Poles have welcomed them—so far.

A massive Russian disinfo campaign is working to change that. washingtonpost.com/technology/202… In 2016, Russian trolls famously used social media to sow division in the US around race, immigration, Hillary etc.

Today, they're trying to do something analogous in Germany, a critical Ukraine ally—tearing apart its consensus on Ukraine from the inside. washingtonpost.com/technology/202…
Nov 9, 2022 • 4 tweets • 2 min read
Facebook is laying off more workers than Twitter ever had. The tech recession is real. washingtonpost.com/technology/202… by @NaomiNixWrites @NaomiNixWrites There's no such thing as a "nice" layoff but worth noting Zuck addressed his staff directly, took responsibility, offered 16+ weeks' severance, immigration help, & a chance to speak w/ someone directly.

Twitter's layoff memo was unsigned and offered close to the legal minimum.
Nov 3, 2022 • 6 tweets • 4 min read
Inside Elon Musk's "free speech" Twitter, a culture of secrecy and fear has taken hold. Managers and employees have been muzzled, Slack channels have gone dark, and workers are turning to anonymous gossip apps to find out basic info about their jobs. washingtonpost.com/technology/202… On Oct. 26, Twitter's head of people emailed staff that Elon was visiting the office, and that they'd all hear from him directly the next day.

Musk took over that night and fired the C-Suite. Staff have not heard from him, or anyone in leadership, since. washingtonpost.com/technology/202… Image
Nov 2, 2022 • 4 tweets • 1 min read
Meeting w/ civil rights groups and punting on Trump was smart on Musk's part. Right now the biggest threat to Twitter's business (of his own making) is that it will be seen as unsafe for marginalized groups and hostile to the left, sparking an exodus. washingtonpost.com/technology/202… As many have observed, Musk is speedrunning the history of online content policy. There's a reason every other successful social platform desperately tries to avoid being seen as having a partisan bent. It's constant triage trying to mollify the side that's madder (or in power).
Sep 7, 2022 • 8 tweets • 3 min read
I reported last month that Twitter would be expanding its Birdwatch fact-checking project after seeing encouraging results: washingtonpost.com/politics/2022/…

Here's Twitter's announcement today about the expansion and a new reputation system to guard against abuse: blog.twitter.com/en_us/topics/p… The extremely obvious worry with crowdsourced fact-checking is that it will be gamed by people with an agenda. Birdwatch's new system will require you to build up a positive reputation by accurately rating others' fact-checking notes (per consensus) before you can write your own. Image
Aug 12, 2022 • 9 tweets • 3 min read
A lot of politics on social media can be understood in terms of supply and demand. There is always heavy demand for information that suggests your side is right and the other is wrong. When the mainstream media don't satisfy that demand, social media will elevate those who do. If you looked only at mainstream media right now, you'd read a lot of fact-based reporting that makes Trump look bad. But if you're a Trump fan, you don't want that. So you look to Twitter for claims that it's all a hoax--and you'll always find folks willing to provide those.
May 20, 2022 • 8 tweets • 3 min read
OK obviously I’m late to this but how wild is it that Insider contacted Elon for comment yesterday morning, and by yesterday afternoon he was publicly announcing that he’s become a Republican and to expect politically motivated “dirty tricks” from the left in retribution?
Apr 27, 2022 • 8 tweets • 4 min read
Elon Musk has now publicly agreed with and amplified criticisms from the right of two individual Twitter employees today— one accusing its top policy exec of “censorship” and the other accusing a company lawyer of facilitating fraud. Imagine getting up and going to work in the morning for a company whose new owner is systematically attacking your colleagues for his 84 million followers, in public, on the site you work for—and knowing you might be next. This is the worst-case scenario Twitter employees feared.
Apr 21, 2022 • 33 tweets • 7 min read
I was apprehensive about Obama taking on "disinfo" but he's starting off on a solid, nuanced note: framing social media as accelerating existing cultural divides, some of which is an inevitable byproduct of a connected world and some the result of specific decisions & biz models. Obama rightly frames social media as part of a broader infosphere, w/ platforms both feeding on and influencing traditional media.

"Tech companies & social media platforms are not the only distributors of toxic information, I promise you: I've spent a lot of time in Washington."
Apr 13, 2022 • 5 tweets • 4 min read
So much for "the first TikTok war." In Russia, TikTok has muzzled its own users and blocked out the entire outside world, creating an alternate universe where pro-Putin propaganda thrives and no one is against the invasion. washingtonpost.com/technology/202… New research from @trackingexposed shows how TikTok's block on new content in Russia successfully silenced anti-war views, even as loopholes allowed pro-war propaganda through. I wrote about their latest findings, released today: washingtonpost.com/technology/202…
Mar 23, 2022 • 6 tweets • 1 min read
in case you were wondering how things are going these days for the generation that attended kindergarten exclusively via zoom...

our 6yo has started surreptitiously playing wordle on his school-issued chromebook while in class and skyping us his score oh also, our 6yo's first-grade class uses online reading software that rewards you with points for completing stories. you can use the points to buy accessories and flair for your avatar. at some point one of them figured out that they all have the same password...
Oct 26, 2021 • 7 tweets • 4 min read
Facebook secretly weighted reaction emojis, including "angry," as 5x the value of "likes"--over the integrity team's warnings.

We wrote about the obscure, often arbitrary, human decisions that shape Facebook's algorithm and how we all interact online: washingtonpost.com/technology/202… To me this is not a story of Facebook intentionally fanning anger for profit. It's a story of how arbitrary initial decisions, set by humans for business reasons, become reified as the status quo, even as evidence mounts that they're fueling harms. washingtonpost.com/technology/202…
Oct 5, 2021 • 4 tweets • 2 min read
This stand is not as incoherent as people are making it out to be. If you view FB's problems as largely common to other social platforms, and you view regulation as the answer, then a fragmented, competitive market arguably makes those problems less tractable than a monopoly. To be clear, I'm not endorsing Haugen's view. I think there are a lot of good arguments for antitrust action to break up the platform monopolies. Just saying it's not crazy and doesn't make her dumb or a techno-libertarian or a Big Tech apologist as some dunkers keep suggesting.
Aug 18, 2021 • 8 tweets • 5 min read
Facebook is launching its own answer to @kevinroose's popular @FacebooksTop10 account. @kevinroose @FacebooksTop10 Facebook has long complained that @FacebooksTop10, dominated by right-wing sources such as Dan Bongino & Daily Wire, represents a skewed picture of what's popular on its platform. I wrote about this debate in depth last year: onezero.medium.com/the-battle-ove…
Aug 17, 2021 • 9 tweets • 5 min read
This is provocative, erudite, historically informed, and worth reading for anyone who thinks about "disinfo."

I think it's also deeply unsatisfying and I have a couple of specific critiques. But the questions it raises are ones that deserve raising. harpers.org/archive/2021/0… Off the bat, @Bernstein absolutely nails the unspoken, unexamined assumptions of a lot of the less-sophisticated hand-wringing over our "broken" modern information sphere. It has never not been broken. harpers.org/archive/2021/0…
Jul 22, 2021 • 4 tweets • 5 min read
Facebook and YouTube spent a year fighting covid misinformation. It's still spreading, @rachelerman & @GerritD report. washingtonpost.com/technology/202… @rachelerman @GerritD How are people evading social media's vaccine misinfo policies? One way is coded language, like calling the vaccine "Maxine" and the dangers of "dancing" with her. One FB group called "Dance Party" has 40k members. nbcnews.com/tech/tech-news… by @BrandyZadrozny & @oneunderscore__