Shirin Ghaffary
@shiringhaffary
Followers
12K
Following
5K
Media
239
Statuses
6K
Reporter at @business covering AI. Send me your AI questions, hopes, fears: [email protected] DM me off-the-record on Signal. username: shirin.30
Joined April 2009
SCOOP: Around 2,300 cafeteria workers who work at dozens of Google campuses in the Bay Area, including the company's headquarters in Mountain View, have unionized.
12
636
3K
36 Amazon warehouse workers in Sacramento walked off the job early Monday morning, demanding paid time off. It's the latest example of Amazon workers who say they're doing back-breaking jobs pushing for better labor conditions. More here:
32
970
2K
NEW: Perplexity closed its latest funding round of $500 million at a $9 bil valuation, led by IVP, tripling its valuation since June
34
132
2K
NEW: OpenAI is preparing to launch a computer using AI agent codenamed “Operator” that take actions on a person’s behalf thru a browser, such as writing code or booking travel. Staff told in all-hands mtg today it will be released in Jan. w @rachelmetz .
68
300
2K
NEW: OpenAI pitched the White House on building data centers in the US as large as 5GW capacity -- for ref, that's enough to power 3 mil homes. OAI's analysis says it could add tens of thousands of jobs, boost GDP, and keep US ahead of China on AI. Link:.
62
129
835
I profiled Chris Smalls (@Shut_downAmazon) as he plans what's next for the ALU, the first union to win the right to represent Amazon warehouse workers in the US. “We’ve got 10,000 people’s lives in our hands" .
4
144
694
Excited to announce I’ll be joining Bloomberg.@business as an AI, ethics and policy reporter starting in Sep. I'll also be starting a weekly newsletter to help readers understand the both thrilling and terrifying world of AI (send me all your burning questions!) (1/3).
Excited to welcome 3 all-stars to the @technology.AI team: @shiringhaffary to cover AI ethics/policy and pen a newsletter that will guide readers into the AI age. @sfiegerman to help edit and pilot our AI coverage. and the incredible @LeonYin on data-driven investigations!.
83
31
663
It's my first official day at @business covering AI, ethics and policy💫. Let's talk! You can now catch me on the terminal or sghaffary@bloomberg.net.
33
24
551
Google’s attempt to shut down a unionization meeting in Switzerland just riled up its employees, who held the meeting anyway.
5
167
464
Excited to announce that going into 2020 I’m starting a new position at Recode as a senior reporter covering social media/big tech. If you work at or have tips about Facebook, Twitter, Google, or other major tech companies, please get in touch! My DMs are open and confidential.
17
15
500
New from me: Students at major universities like UC Berkeley, Brown, and Yale are fighting to keep Palantir off campus over the company’s contracts with ICE. And they seem to be getting some results.
2
141
328
A group of contractors at Google who hold college degrees, get paid as little as $40,000 a year, and don’t receive sick days are voting to unionize today. I wrote about why the vote is important and what it means.
2
86
327
NEW:. After months of private meetings with Facebook, Democratic Party leaders say the company is unprepared to handle election misinformation. "They’re essentially slapping Band-Aids on wounds that require emergency room stitches,” said one source.".
20
142
295
ok it's official: Meta is designing legs for avatars. full body avatars expected to come out next year, Zuck says Meta will use predictive AI to build it.
68
8
320
West coast/east coast twitter divide very evident right now between ppl freaking out about OpenAI v Eric Adams. Remarkably little overlap.
18
28
328
The "I did not leave my country for this" line is such an effective guilt tripping tool in any immigrant parent's arsenal. Not sure I will ever be able to counter that.
My dad later today (not kidding). "Yashar, I did not leave Iran and move to America so my son could tweet about towels. What is this nonsense? If you were my student I would give you a D minus. I don't have a son, I have a silly goose!".
15
30
281
NEW: Meta CMO told employees not to get too excited about recent stock jump, according to an internal memo from earlier this month. The company is "still at the whim of Apple," making progress but "not satisfied" with number of younger users in US.
1
56
96
Hearing about Twitter employees already being locked out of their work emails tonight, and receiving no formal email/notice if they're actually laid off.
11
108
274
The world’s biggest tech conference for women, just dropped Palantir as a sponsor over its controversial work for ICE. A VP from @AnitaB_org stated that the company's tech "enables the human rights abuses of asylum seekers" .
2
84
248
statement from Google union on layoffs today:. "With billions in profits and executive compensation untouched, our jobs should not be on the chopping block."
5
75
249
New letter from 2 former OpenAI employees who have been advocating for whistleblower protections in AI on SB 1047 legislation:. “Sam Altman, our former boss, has repeatedly called for AI regulation. Now, when actual regulation is on the table, he opposes it.”
23
62
266
SCOOP: OpenAI is opposing controversial AI safety bill SB 1047, arguing that it would "slow the pace of innovation", and bc of nat'l security implications, AI should be regulated on a federal rather than state level.
25
41
265
Amazon has fired several black employee organizers in the past few months who have been pushing for better working conditions during the pandemic (it claims for unrelated reasons). The company never apologized to Chris Smalls, who its top lawyer called "not smart, or articulate".
7
115
233
NEW: Per internal memo, OpenAI is still involved in "intense discussions" with the board, Sam Altman, Emmett Shear to unify company. No resolution this evening. Discussions will resume tmrw morning “when everyone’s had a little more sleep,” .
18
43
243
NEW: Liquid AI, a startup designing AI inspired by the structure of the tiny brain of a worm, is set to raise $250m in funding at a $2.3 bil valuation 🪱. Liquid's models are a promising + more efficient alternative to transformer models, investors say.
15
44
251
New from me: . Google says it’s committed to ethical AI research. But several researchers I spoke with on its ethical AI team aren't so sure.
10
57
216
NEW: We found several Facebook militia groups + pages in which members openly encouraged violence, including shooting BLM protesters. The company took down at least four of those pages down after we flagged. Tip @Techmeme .
13
103
185
New OpenAI CEO.
Unironically. Most of the CEO job (and the majority of most executive jobs) are very automatable. There are of course the occasional key decisions you can’t replace.
11
41
198
Twitter just locked Trump's account for 12 hrs. Our story on how social media platforms handled the President's encouragement of today's violence. And the increasing calls for Twitter/Facebook to deplatform him altogether.
6
51
196
NEW: In Mark Zuckerberg's leaked internal video call with employees this morning, he said Trump's "looting. shooting" comment about protesters in Minnesota “has no history of being read as a dog whistle," for inciting vigilante violence. More here: .
22
104
179
Hearing that employees at OpenAI are stunned by the news about OpenAI CTO Mira Murati’s departure, many replied with a “wtf” emoji to an internal note announcing the exit .
14
32
206
NEW: The pandemic has become a flashpoint in the battle over how Amazon treats its workers. I spent the past 3 months investigating what’s happening at Amazon’s warehouses, how that affects everyone else, and how we got here in the first place w/ @delrey .
4
80
192
NEW: The NLRB added three more fired Google workers to its growing complaint against the company for allegedly retaliating against employee activism. The case could end up changing the scope of what you're legally allowed to say at work.
1
64
167
A new study shows that AI models are 1.5 times more likely to flag tweets written by black authors as “offensive” compared to other tweets. I wrote about why that is and why it's a problem:.
19
77
163
Trump has been building up to this moment for months by perpetually promoting the conspiracy theory that the election would be "rigged" against him, before it even started. I wrote this piece prior to election night, but it's even more relevant now.
2
21
161
At the Facebook whistleblower Senate hearing today. Will thread live tweets here.
3
27
158
Over 200 employees at Tableau protested the data visualization company’s work with ICE and CBP today over human rights concerns— joining similar calls from workers at Amazon, Salesforce (which recently acquired Tableau), Github, Microsoft, and other tech cos.
Today we say #DrawALineTableau and call on our company to publicly cease its policy of complicity towards the systematic human rights abuses being carried out by CBP and ICE!. Here is our statement.
0
45
141
For this week's newsletter, I wrote about the OpenAI rumor mill, and had a really fun time explaining "Strawberry Man" and "Jimmy Apples" to my editors in the process🍓.
18
11
166
New from me: Around 200 Google employees and protesters rallied outside the company's San Francisco offices today, protesting the company's suspension of two employees who were involved in workplace activism.
5
48
147
Major news in CA: . Gov Newsom has vetoed AI safety bill 1047, which has divided the tech community in recent months. In his veto message, Newsom said he still believes AI risk important to mitigate, but disagreed with specifics of the bill.
7
24
165
I asked a dozen leading voices in tech, politics, and civil society how they would fix Facebook. Their solutions range from breaking up the company, to to creating a new FDA-like agency just for social media.
10
49
139
If OpenAI's board members have substantial safety concerns about how the company was being run, now would be the time to elaborate.
1
29
128
New from me: Dozens of Google employees say they were retaliated against for reporting workplace issues like sexual harassment. We reviewed internal docs + talked to Googlers about the cont’d issues for those who speak up, nearly a year after the walkout.
4
87
117
NEW: Google is starting to warn people when search results on a trending topic are new and potentially unreliable. The company confirmed it starting testing this feature about a week ago. More details here:. and h/t to @noUpside for tweeting about this.
3
44
124
NEW from me and @DelRey: An internal Amazon memo reveals plans to build a sophisticated software tool to track unionization attempts in its workforce, among other perceived threats to the company's operations like local weather + crime.
9
93
119
I interviewed the "godmother" of AI @drfeifei ahead of the release of her new memoir. We talked about why she thinks the US needs to take a "moonshot mentality," to funding AI research, and why she isn't an AI doomer like some of her academic peers.
6
33
129
Google employees protesting YouTube's harassment policies at SF Pride Parade today. About 23 in the group. @NoPrideForGoog
2
35
116
Amazon employees shared their shock & anger with us about a top exec targeting a fired warehouse worker who spoke out about coronavirus safety concerns on the job. If you work at Amazon or in tech and have thoughts on this/similar situations, pls DM me.
0
42
108
I wrote about how Twitter's quality as a product is degrading a month ago. Blocking Substack links is another example of that — not to mention an obvious contradiction of Musk's stated freedom of speech ideology.
4
41
117
NEW: Former OpenAI board member Helen Toner says that when ChatGPT came out, the board found out about its release from Twitter, and wasn't informed prior to launch. One of several new details that Toner shared on a recent TED podcast. Story here:.
10
26
127
Zatko says that the Indian foreign agent working for Twitter wasn't an engineer (who could presumably look up company data), and was hired to be an "inside" negotiator w/ Indian gov't. Sen. Lee asks him whether the Indian agent had "access" to engineers, he says yes.
3
76
122
My latest and last (🥲) article for Vox is about the growing open v. closed source debate in the AI world. There's a clear biz story here but also an ethical one: is it more dangerous to keep AI models in the hands of a few, or share it with the masses?.
4
24
125
This seems like a labor complaint waiting to happen. Under the NLRA, it's illegal for employers to prevent its workers from discussing wages and benefits.
NEW: Apple employees were just barred from creating a Slack channel to discuss pay equity. Here's what Apple HR had to say about it: 1/.
3
27
113
Very excited to finally announce what I’ve been working on for the past few months: hosting the next season of Vox's Land of the Giants podcast about Google with @Kantrowitz. Trailer is out today, first episode comes out next Tue. Listen below!.
6
24
120
Excited to announce that I'm now a senior correspondent for Vox and Recode, covering the intersection of social media, politics, and policy. If you work in social media (esp. at FB/Google) and want to talk, please get in touch: shirin.ghaffary@protonmail.com, DM for Signal (1/2).
13
7
120
For some reason Google thinks I’m @linakhanFTC. Haven’t written an article about her in a while so really curious why this is happening.
13
5
117
“Don’t be evil” isn’t a normal company value. But Google isn’t a normal company. In the 1st episode of the new season of the Land of the Giants podcast, @Kantrowitz and I tell the story of how Google adopted the mantra 20+ yrs ago and why it still matters.
7
33
108
NEW: Elon Musk just sent the following note to Twitter staff:. “I cannot emphasize enough that Twitter will do whatever it takes to adhere to both the letter and spirit of the FTC decree. Anything you read the contrary is false" (1/2).
2
9
105
“Last week in advance of our hearing Facebook released two studies and said that the Wall Street Journal was all wrong,” said Sen. @MarshaBlackburn "Having seen the data, I feel pretty confident that Facebook has done the misrepresenting to this committee.”.
3
23
105
Wrote an EOY piece about the growing importance of tech whistleblowers and activists in holding the industry accountable for its problems. Thanks to everyone who spoke to me about their personal experiences coming forward.
2
35
100
NEW: OpenAI says it is releasing a majority of past employees from nondisparagement agreements tied to their exit contracts and equity . Gift link:.w/ @rachelmetz
4
28
107
Bloomberg's new weekly AI newsletter, Q&AI, launches TODAY. @BradStone is kicking off our first edition (I've read it — it's a good one) and I'll be taking the reins after that. Sign up here. It's free!.
4
15
109
Very excited for this!!! Sign up for my *free* AI newsletter coming to your inboxes shortly. Link below.
Are AI stocks for real? Can ChatGPT do my job? What is Stable Diffusion? . You’ve got questions. We’ve got answers. Introducing Q&AI, a new weekly newsletter from Bloomberg Technology’s @shiringhaffary
6
13
102
Sam Altman told employees today that OpenAI is on track to take in $3.4 bil in annual revenue -- about $3.2 billion from its own revenue and $200 mil thru its share of Azure API rev. props to The Info for the scoop, which our reporting also confirmed.
6
20
106
Both @robinw and @yoyoel were on a call yesterday reassuring advertisers that Twitter will remain a safe place for brands. Also some of the few remaining public/client facing execs left at the company from before Elon times.
Robin Wheeler, VP of US client solutions, is out as well, per sources.
5
36
78
Met w/ @donttrythis today to talk about the March for Science. More to come in our next episode! #kqedabovethenoise
0
6
96
Talked to over a dozen leading social media researchers for my latest:. Why no one really knows how bad Facebook’s vaccine misinformation problem is.
4
50
94
Trump used to be everywhere on social media and now it feels like he's nowhere. I wanted to know just how much ppl stopped talking about him since he lost the election and was banned from Twitter + Facebook. Here's what @ranimolla and I found .
8
36
92
Sergey Brin is roaming around I/O; he answered reporters’ off the cuff questions for a few minutes. Story incoming!
2
4
100
Mozilla says it's done an independent review of the code that NYU Ad Observer was using to collect opt-in user data, and says that Facebook's privacy concerns aren't valid.
Facebook claims they shut down the researchers accounts due to privacy problems with the Ad Observer. In our view, those claims simply do not hold water. We know this because we reviewed the code ourselves before recommending it. Read more here:
0
39
88
My first weekly newsletter for Bloomberg is out!. I wrote about what to make of the star-studded, closed-door AI summit in DC yesterday, and a Q&A with Hugging Face CEO @ClementDelangue, who was in the room for it all. Sign up here (it's free):.
3
18
89
Another big development in Google’s apparent upheaval of its ethical AI team. Meg Mitchell, who used to co-lead the team with Timnit Gebru, says she’s been fired. We’ll explain this situation in detail and why it’s so controversial in an upcoming episode of our podcast.
I'm fired.
5
30
92
Sen @MarshaBlackburn asks why Facebook allowed sex trafficking on its platform until Apple threatened to kick it out of the app store. Davis: "Respectfully I do not agree. ". Sen. Blackburn: "Miss Davis, this is your company's reporting. You knew this was there.".
4
34
91
An emerging line of defense by FB against Haugen's testimony today: she didn't work on all the projects she's taking issue with. That's true, but she also revealed thousands of docs which back up her claims. She says these docs were accessible to anyone working at Facebook.
Just pointing out the fact that @FrancesHaugen did not work on child safety or Instagram or research these issues and has no direct knowledge of the topic from her work at Facebook.
4
31
73
Pre-Christmas SCOOP:. OpenAI Is in early talks to raise new funding at a valuation at or above $100 bil, according to people with knowledge of the matter. w/ @EdLudlow @GillianTan . With that, I'm out for 2023. It's been a wild one.
6
21
93
This is the fourth tweet by Trump in recent weeks that Twitter has marked up for violating its policies around things like voter misinformation, manipulated media, and in this case, abusive behavior. Here's our previous coverage on the topic.
3
35
85
Elizabeth Warren, who has made breaking up Big Tech a central tenant of her presidential platform, is the most funded candidate by Google employees. @ranimolla and I took a close look at why that is:
4
46
82
Atop the @salesforce tower and ready to roll for episode 3 of Recode/MSNBC #Revolution with @karaswisher and @Benioff
4
10
90
NEW: Perplexity is the latest tech company donating $1 million to Trump’s inaugural fund, joining Sam Altman, Meta, Amazon who did the same.
8
22
92
It was a treat talking to the always clear-eyed @RebeccaSolnit about the value of writing in the time of AI . "I often feel that I don't think hard enough about things until I have to write about them."
3
21
82
Protest outside of Palantir HQ in Palo Alto today over the company’s multimillion dollar contract with ICE. Chanting things like “Palantir, go away, how many kids did you deport today?” Organized by @CloseTheCampsBA coalition
1
35
73
NEW: Anthropic is releasing an agentic tech for developers that can click, type, and browse the web on a user's behalf by analyzing what's happening on their screen. Called "computer use." More here:.
1
26
89
NEW: Anthropic CEO Dario Amodei called for mandatory AI safety testing when speaking at a gov’t conf today:. “I think just public attention and the fact that employees care has created some pressure, but I do ultimately think it won’t be enough.”.
4
23
90
OpenAI's head of alignment, John Schulman, is leaving for Anthropic. Schulman is a co-founder, and well-respected w/in the company, so this is a big deal.
I shared the following note with my OpenAI colleagues today:. I've made the difficult decision to leave OpenAI. This choice stems from my desire to deepen my focus on AI alignment, and to start a new chapter of my career where I can return to hands-on technical work. I've decided.
2
9
89
NEW: Google employees are holding a public rally outside its SF offices at 11 am Friday in solidarity with two coworkers who have been put on admin leave for looking up internal docs and allegedly violating company policies -- calling it an attack on transparency + accountability.
2
25
78
NEW:. For @BW, I profiled OpenAI's quiet dealmaker, COO @BradLightcap. The 33-year-old is leading critical negotiations with major media pubs, and a whole lot more as OpenAI is expanding its business model. Gift link below:. edited by @joshuabrustein.
4
31
83
Talking about Anthropic's new release w/ @EdLudlow on tv today + debuting the fab earrings my 8 year old niece made for me :)
5
3
86
More news from Facebook today: the company is suspending all political ads after the polls close, indefinitely + banning "militarized language" around poll watching. In the future, Donald Trump Jr's recent "Army for Trump" video would be taken down under new policy.
10
19
66
Google is trying to build the most futuristic technologies in the world in its moonshot factory. This week we tell the story of how those projects can succeed, how they can fail, and what it means for the rest of us. Listen here:.
4
23
77
This week on Land of the Giants:. We talked to Dr. @timnitGebru, Google's former co-lead of its ethical AI team, whose controversial exit caused an outcry and called into question whether the company was trying to silence critical research (1/5).
3
35
77
side-note: @ruchowdh managed a team doing important research at Twitter around algorithmic fairnesss and transparency. Last year, the team put out a report showing how Twitter's algorithm often amplifies right-wing over left-wing political content.
4
20
71