Now more than ever trying to live by these rules:
📍Show Up
📍Pay Attention
📍Tell the Truth
📍Don’t be Attached to the Consequences
Pinned Tweet
16
45
239
Follow
Click to Follow JuliaAngwin
Julia Angwin
@JuliaAngwin
Investigative journalist. Founder of . Formerly , . Author of Dragnet Nation, Stealing MySpace. I'm more active at @julia@journa.host.
Julia Angwin’s Tweets
In a conversation with about AI hype, I argue that the most worrisome uses of ChatGPT aren't malicious ones but rather everyday people and organizations using it to cut corners due to the pressures they face, like CNET's misguided experiment.
5
78
237
Show this thread
AI also should not be deciding the line between acceptable and unacceptable speech, says. It's messy, it needs input from civil society, it's constantly shifting and culture specific.
Because of that, he says, "AI has no role here." /end
3
21
Show this thread
Other kinds of AI are far more dangerous, says, particularly ones that aim to predict social outcomes -- such as in hiring and criminal risk predictions.
These types of AI should be considered "illegitimate and morally impermissible." /4
1
15
39
Show this thread
However, doesn't believe that ChatGPT creates a doomsday scenario that destroys jobs and harms the economy - as some have warned.
Even larger innovations such as the internet and smartphones haven't done that. This is not a "sky is falling" scenario, he says. /3
1
4
13
Show this thread
ChatGPT's tenuous relationship with the truth means that it can be a dangerous creator of misinformation.
Add economic pressures that force institutions to cut costs and rely on automated content creation and you have "a recipe for disaster" says . /2
2
9
34
Show this thread
ChatGPT is great at its job. But its job is to be a bullshit generator -- an automated creator of plausible lies.
"It is very good at being persuasive, but it’s not trained to produce true statements" says in my latest newsletter. /1
22
185
487
Show this thread
There is a lot of work to be done to counteract rising political violence and the speech that incites it.
But first we have to be clear about the fact that incitement is often based on fear and does not always require hate. My newsletter here:
6
18
Show this thread
Tech platforms could also do more by directing less attention to controversial but not dangerous hate speech and focusing instead on truly dangerous speech, recommends.
1
1
14
Show this thread
What is the solution to dangerous speech? Counterspeech. "Influential people of various spheres need to refrain from dangerous speech themselves and denounce it when other influential people use it," says.
1
2
7
Show this thread
"In the United States at the moment, there is, at minimum, a striking and alarming shift in the extent to which dangerous speech is used and condoned by political leaders and other influential people," told me.
2
1
3
Show this thread
This kind of dangerous fear-mongering is on the rise in the United States. Think of the myth of the "great replacement" which claims there is a conspiracy to replace White people - a myth that prompted violence.
1
7
7
Show this thread
"I was stunned at how similar this rhetoric is from case to case. It’s as if there’s some horrible school that they all attend," told me.
1
2
6
Show this thread
Prior to the Rwandan genocide in 1994, Hutu politicians warned the Hutus that they were about to be exterminated by Tutsi cockroaches.
Prior to the Holocaust, Nazi propagandists declared that Jews were planning to annihilate the German people.
1
2
Show this thread
It is fear, more than hate speech, that often leads to mass violence. Leaders who seek to incite violence often create fake threats so that people will feel they must defend themselves.
1
5
10
Show this thread
But missing from the discussion has been a true understanding of what type of speech incites violence. has been studying the types of speech that have led to genocide, and she has found a consistent pattern.
1
1
4
Show this thread
This week, new details emerged about how social media platforms allowed violent rhetoric to circulate freely on their platforms in the weeks leading up to the Jan. 6 insurrection.
1
9
15
Show this thread
What is dangerous speech and what can platforms do to moderate it in the wake of its rise? ’s turns to , who coined the term, for answers:
7
14
Their working conditions reveal a darker side to the AI boom: that AI often relies on hidden, low-paid human workers who remain on the margins even as their work contributes to a multibillion-dollar industry. (6/8)
15
1,199
3,864
Show this thread
She hopes FB users will eventually be offered a yes/no option for ad profiling. Of course, even then, we could just end up with more annoying take-it-or-leave-it pop-up boxes that are not a true choice.
As is always true with consent, it must be freely given to count. /end
21
112
Show this thread
Frustrated activists are now seeking to go around the Irish regulator. Last month, filed a lawsuit against Meta in the UK demanding that it comply with her denial of consent to profiling.
My interview with her in today’s newsletter /7
1
40
112
Show this thread
Meta has said that it will appeal and doesn’t need to change anything about its ad targeting practices. /6
2
21
72
Show this thread
So the landmark case decided by the EU—which was brought by brilliant strategic privacy litigator in his Forced Consent project—is now back in the hands of for a resolution. /5
1
26
91
Show this thread
But Ireland’s economy is heavily dependent on Big Tech. And its regulator has a backlog of GDPR cases that have waited years for resolution.
As activist tells me, “There is just never going to be political will to properly enforce GDPR in Ireland.” /4
1
42
145
Show this thread
The problem is that GDPR enforcement is not in the hands of the EU, but in the member states. And most Big Tech cases land at the Irish regulator because that is where many tech European headquarters are based. /3
1
37
119
Show this thread
Earlier this month fined Meta €390 million for not getting proper consent before profiling FB & IG users. It was hailed a huge victory for EU’s landmark privacy law, GDPR, but sadly it may not change how you are profiled. /2
2
39
180
Show this thread
Let’s talk about consent. Do you feel like you ever properly consented to being surveilled online constantly, having a profile built of your interests and having that profile made available to anyone who could pay for it?
EU regulators don’t think so either. /1
17
379
1,345
Show this thread
"I don’t think we have much of a chance for a livable future if we don’t have a truly private means to communicate with each other," president tells me in this week's newsletter:
4
33
84
Be hardcore. Support journalism.
Quote Tweet
All the more reason to keep reading Ryan Mac, Donie O'Sullivan, Drew Harwell, Matt Binder, Micah Lee, Aaron Rupar, Keith Olbermann on their web sites.
Subscribe & Support journalism.
Keep accountability reporting alive
7
29
In the end it is always privacy for the rich and surveillance for the rest of us.
twitter.com/osinttechnical
read image description
ALT
read image description
ALT
Quote Tweet
SCOOP w/ @CaseyNewton: Twitter is working on a plan to force users to opt-in to personalized ads & share their location data. It’s considering letting those who pay for Twitter Blue to opt-out of data sharing — a decision that would likely anger Apple: platformer.news/p/twitters-ris
4
69
109
Entirely my pleasure! Loved working with you and will always be your biggest fan 🎉❤️
1
3
🚨Legend alert🚨 It was an honor to work with - an incredible reporter with unmatched commitment to exposing harms against the most vulnerable.
Her series Working for an Algorithm sadly gets more relevant by the day. themarkup.org/series/working
Quote Tweet
Today is my last day at @themarkup, where I can’t believe I was lucky enough to work for two years.
I got to learn from a massively talented crew and was encouraged to write stories that highlighted worker voices. I’m so grateful.
(I’ll say what’s coming next in the new year)
Show this thread
1
1
13
My prediction for journalism in 2023: democracies will get serious about saving journalism.
No pressure 😂
read image description
ALT
4
5
19
📰Video: Dec. 8 talk with (), Apar Gupta (), () and Moderator: discussing investigative news, digital evidence, ethics and The Wire’s expose of Meta's content policies.
🔗Watch: bit.ly/3BoUaRm
5
1
“My avatars were cartoonishly pornified, while my male colleagues got to be astronauts, explorers, and inventors.” 😢
Quote Tweet
I tried the viral Lensa AI portrait app, and got lots and lots of nudes. I know AI image generation models are full of sexist and racist biases, but this one really hit home. My latest story for @techreview technologyreview.com/2022/12/12/106
1
11
Huge honor to speak to for the Markup, as her series of surveillance stories for the Wall Street Journal in early 2010s: “What they know” was a huge inspiration for and I while investigating our own stories on China’s surveillance state.
Quote Tweet
"China’s surveillance state is like a panopticon, where the idea that you’re being tracked is more effective than the actual tracking mechanisms," @lizalinwsj tells me in this week's newsletter with @joshchin:
themarkup.org/newsletter/hel
Show this thread
4
20
It's important to understand China's ever-expanding surveillance regime because oppressive governments love to copy techniques from each other.
1
6
Show this thread
China leads the world in computer vision and other surveillance-focused areas of AI, tells me in this week's newsletter.
In Xinjiang, AI us being used to flag intellectual, cultural, and religious "threats" to the Communist party.
2
1
3
Show this thread
Social control is the entire point of surveillance. So it doesn't matter *too much* how well it works. It matters more how afraid people are of being caught.
That's why AI is a great surveillance tool. It's not always accurate but it can be good enough to scare people.
1
10
Show this thread