Sarah Shoker (@SarahShoker) header
Sarah Shoker (@SarahShoker) profile picture

Senior Research Scholar in AI at the Berkeley Risk and Security Lab at UCB. Formerly the Lead of the Geopolitics Team at OpenAI.

San Francisco, CA
Joined July 2010
Photo by Sarah Shoker (@SarahShoker)Photo by Sarah Shoker (@SarahShoker)Photo by Sarah Shoker (@SarahShoker)Photo by Sarah Shoker (@SarahShoker)Photo by Sarah Shoker (@SarahShoker)Photo by Sarah Shoker (@SarahShoker)Photo by Sarah Shoker (@SarahShoker)Photo by Sarah Shoker (@SarahShoker)Photo by Sarah Shoker (@SarahShoker)Photo by Sarah Shoker (@SarahShoker)
Sarah Shoker retweeted
Absolutely incredible news that Egyptian author @El_Gendy_95, who I am proud to call a friend, has won the 2026 National Book Awards for Nonfiction! Huna is an unforgettable, painful, and brilliant story of political imprisonment in Egypt and a coming of age unlike no other.
19
68
3,259
Sarah Shoker retweeted
Replying to @DKokotajlo
At long last, we will wear the ring of power from the classic novel don’t wear the ring of power
1
8
224
This interpretation isn't in the source material. A council pressures him into the quest. Frodo tries to give away the ring several times & he wears it b/c he falls to temptation. This emergency powers thing is...no. The ring isn't the reasonable limits clause in a constitution.
I also take issue with @slatestarcodex’s use of the metaphor here. Yes Frodo used the ring to hide a few times, all while working to destroy it. AI companies racing to RSI are a lot more like Boromir than Frodo
1
1
288
I'm not here to argue about RSI, but no one's going to malign my boy Tolkien (RIP.)
73
Okay so there's this character called Boromir and you will not believe what I'm about to tell you but
This is not the correct takeaway from Lord of the Rings
1
20
660
Sarah Shoker retweeted
DeepMind engineer Sofia Liguori just resigned from Google in protest of the company's Pentagon contract. I interviewed her about her decision and the future of worker dissent: theintercept.com/2026/10/21/…
1
16
35
2,102
it’s crazy that the world is ending and this is the backdrop to it all
6
90
7,981
Social scientists are told to think of alternative explanations for phenomena that appear to have an obvious cause, so as to offset our own biases. The WSJ story is only a partial view at the facts. Consider instead that OAI leadership has a kink for being hated by the public.
Reporting from the WSJ. That fact that three alignment/safety researchers left this morning was already being discussed here on X.
1
1
21
1,933
Sarah Shoker retweeted
AI exceptionalism has led to a total failure to uphold our standards for what products are deemed safe for use, directly causing harms that are catastrophic to those experiencing them and that in many cases cannot ever be remediated once they occur.
1
7
9
826
Did this actually happen? I authored work on algorithmic bias *in* autonomous weapons as early as 2018. Lawyers and political scientists have been on board for 10+ years. My experience is that AI companies were the ones who were behind here.
old enough to remember when autonomous weapons were a tactic to avoid discussing actual harms like algorithmic bias
1
3
20
1,498
Sarah Shoker retweeted
I resigned from Google DeepMind bc it broke its founding promise by selling AI to the military without restrictions against killer robots or mass spying. For months, I worked to stop this but watched powerful ethicists and institutions choose silence. Here's what happened. 🧵
347
4,729
281
18,174
1,106,440
Sarah Shoker retweeted
We cannot consider #AI to be morally neutral. In reality, every technical tool embodies choices and priorities through what it measures, ignores, and optimizes, and how it classifies people and situations. Ethical discernment cannot be limited to asking whether we are using a system for good or bad purposes. It must also examine how that system is designed and what vision of the human person and society is embedded in the data and models that guide it. #MagnificaHumanitas
1,379
35,736
2,800
213,357
10,577,950
Sarah Shoker retweeted
A good time to remind people that in my time doing LLM research I feel like a minority of my colleagues are American citizens. It would be industry destroying to have to rebuild with segregation for frontier ai research to be legal.
39
62
7
958
57,961
Sarah Shoker retweeted
The US military has backed AI development for decades; in turn, AI has transformed warfare. Amid shifting alliances, lawsuits & melodrama, it’s easy to lose sight of the larger context: In recent years, AI has enabled more/faster killings than ever before. theverge.com/ai-artificial-i…
48
4
109
9,847
Sarah Shoker retweeted
Before Stuxnet, there was fast16. Our team uncovered a 2005 sabotage framework that used a boot-start filesystem driver to silently patch high-precision calculation software in memory. Nudging calculations just enough to throw them off. sentinelone.com/labs/fast16-…
3
44
4
183
34,151
Sarah Shoker retweeted
new @NewYorker column: the AI industry's catastrophic messaging problems, the paradoxes of commercial hype and real-world danger
16
39
8
276
93,161