Opens profile photo
Follow
Emily Baker-White
@ebakerwhite
tech reporter . Formerly , policy for , before that. ebakerwhite@forbes.com, emilybakerwhite@protonmail.com, DM for Signal
San Francisco, CAJoined November 2016

Emily Baker-White’s Tweets

What an entirely lovely thread.
Quote Tweet
THREAD: I spent the long weekend in a house full of formerly incarcerated TikTokers Between us, we did ~70 years & have >5 million followers. A few years ago none of us would have thought this possible Here’s a 🧵w/some of the cool things my friends are doing with their freedom
Show this thread
Image
1
In a dramatic reversal, Cloudflare has dropped KiwiFarms. CF insists its policy is still the same and that it was still wrong to drop 8chan & the Daily Stormer (!), but says KF is different because it’s causing a unique, imminent threat to human life.
1
2
Show this thread
Interesting that Biden might want an EO, in addition to the to-be-finalized CFIUS contract, to be part of the picture. Obviously the devil will be in the details, though: will a video taken by an Indian citizen of their US citizen friends while on vacation in Europe be in scope?
1
Show this thread
Biden is reportedly considering an executive order that would limit data collection about American citizens by Chinese-owned companies. Sounds like a real headache for the companies, given that users hardly declare their citizenship when they sign up for an app.
Quote Tweet
Interesting scoop from @ReedAlbergotti (and lol of course @semaforben would scoop his own launch) medium.com/semafor-media/
1
4
Show this thread

Topics to follow

Sign up to get Tweets about the Topics you follow in your Home timeline.

Carousel

I highly recommend this talk about the data companies might use to create realistic VR experiences. Do we really want to trust big tech companies with these unique physical indicators just so they can ...build us cooler videogames?
Show this thread
Looks like ByteDance's latest VR headset, like Meta's, will track users' eyes and facial movements. Companies use our biometrics to make VR experiences more convincing. But should we really trust them with personal info like what raises our heart rate or makes our eyes dilate?
Quote Tweet
New details about ByteDance’s next Pico VR headset emerge ...Including Mixed Reality features protocol.com/bulletins/pico
1
2
Show this thread
It then says: "We don't and won't talk about these efforts publicly because we don't do them for marketing purposes; we do them because they are aligned with what we believe is morally correct." Ok, but, uh, it's it in your company blog post.
1
5
Show this thread
In a blog post, Cloudflare says it was wrong to drop 8chan and The Daily Stormer as customers. When it provides services to morally abhorrent orgs, the company says it will show its values by donating to groups that oppose those abhorrent orgs, sort of like a carbon offset.
1
7
Show this thread
This has been a common theme worldwide with government bodies seeking to rein in big tech. Regulators have gotten much more sophisticated in a short period of time. But "Senator, we run ads" was just 4 years ago. youtube.com/watch?v=n2H8wx
Quote Tweet
ByteDance went to CAC to discuss their algorithm filings, WSJ reports. “officials at the agency displayed little understanding of the technical details and [they] had to rely on a mix of metaphors and simplified language to explain how the recommendation algorithm worked” twitter.com/_KarenHao/stat…
Show this thread
1
This is a big reason why ByteDance and TikTok are trying to separate their algos. But one thing I learned listening to all those Project Texas recordings: whether an algo belongs to BD or TT is messy at best. And many TT engineers are BD employees reporting to BD leadership!!
Quote Tweet
Governments around the world are grappling with how to stamp out misinfo & toxicity on the internet. China is first to advance its solution: an unprecedented plan to control platforms' underlying algorithms. But there's a problem. It may be impossible. wsj.com/articles/china
Show this thread
5
It, er, says a lot about a candidate's dedication to transparency when they block their public campaign website from being archived.
Quote Tweet
Replying to @KFILE @Bencjacobs and @katherinemiller
He requested his site blocked by @internetarchive twitter.com/z3dster/status
6
CatDossier.docx!!! Hoping Luisa is safe and happy wherever she is now.
Quote Tweet
But what about the cat, many ask. What was Luisa's role? Was that even her real name? Well, built a full-fledged cat dossier for Luisa. Our hope was to find her owner's identity via her cat - "the only stable thing in her life".
Show this thread
Image
Image
2
ByteDance has launched a search engine in China. The "stickying" of propaganda at the top of the new app sounds very similar to what former employees described ByteDance doing with pro-China content at its former US news app, TopBuzz. buzzfeednews.com/article/emilyb
Quote Tweet
TikTok owner ByteDance has quietly launched a search app in China ... where Google is banned scmp.com/tech/big-tech/
5
Ranking changes (boosts, demotions) are moderation. Recs policies too. Relevancy determinations too. They’re all choices about what to show you, and how much of it, and in what order, made by companies trying to maximize profit (and avoid regulation that will impede that profit).
Quote Tweet
Excited to share my newest article, in @SocialMedia_Soc. “Do not recommend? Reduction as a form of content moderation”. I’ve been yakking away about this for a while now, so let me just say what I think really matters here: journals.sagepub.com/doi/full/10.11 A 🧵:
Show this thread
3
seems fine
Quote Tweet
In 2020, ex-WeWork CEO Adam Neumann invested in a women-led startup called Alfred. Less than two years later, his new company Flow – backed by $350 million – sounds strikingly similar. @_IainMartin and I spoke to 8 sources to investigate for @Forbes: forbes.com/sites/iainmart
Show this thread
3
San Francisco: read. this. piece.
Quote Tweet
“If the rising fears among property owners are reflective of anything, it’s the rising desperation among the dwindling pool of locals holding on to a city that continues to drive out its most marginalized communities.”Wrote abt San Francisco’s crime panic: buzzfeednews.com/article/albert
3
Deleted a tweet about TikTok’s in-app browser that might’ve been misleading. The browser contains code that *can collect* all a user’s keystrokes, including passwords and other sensitive info. But the company says they’re not currently using it – at least not that way.
Quote Tweet
NEW: TikTok's in-app browser injects code that could let the company monitor a user's keystrokes and taps on outside websites, according to research by @KrauseFx forbes.com/sites/richardn
Show this thread
1
4
Show this thread
Quote Tweet
NEW: Following a @Forbes investigation, bipartisan Senate leaders are demanding answers from TikTok and its outside content moderator, Teleperformance, on their handling of child sexual abuse material. forbes.com/sites/alexandr
Show this thread
1
7
An easy fix: expand the ad library to show all ads, current and former. This would enable us to track other shady advertisers — scammers, etc. — better, too! And then if I want to see Dinesh's total spend, I'll know it's right and includes even the ads that avoided detection.
Show this thread
Anyway: ad libraries are good, more platforms should have them, and they should focus on more than political ads! But: when your transparency tool only shows things you categorize as X, and your categorization is, er, error-prone, that might not be a great transparency tool.
1
Show this thread
This ad wouldn't have stayed in the Ad Library past its live run, and wouldn't have factored into other Ad Library tools (like its new Spending Tracker), if we hadn't asked FB about it. There are a lot of others like it, enough that I start to worry those tools aren't reliable!
1
Show this thread
So, unfortunately, it appears advertisers *are* avoiding disclaimers by making their ads about merch. Here's one from noted conservative provocateur Dinesh D'Souza. We shared it and some others w/FB, which is no longer letting it run without a disclaimer.
Image
1
Show this thread
FB uses a 'primary purpose' test to determine whether an ad is political. This is good for edge cases; advertisers can't avoid disclaimers by just, say, slapping their text on a piece of merch. But you can't teach an ML model a 'primary purpose' test. It scales *horrendously.*
1
1
Show this thread
FB's Ad Library is a big part of their transparency strategy, but it doesn't treat all ads equally. It shows all *current* ads, and then keeps *certain* other ads (incl political ones) after they've run. When FB errs in categorizing ads, the Ad Library's reliability suffers.
1
Show this thread
Update: After this story was published, Flow's job board disappeared from the website. Adam Neumann's company had been seeking web3 engineers who will help “redefine the economic experience of renters” by “leveraging new technology.”
15
60