Hannah Bailey (@Hannah_LSBailey) header
Hannah Bailey (@Hannah_LSBailey) profile picture

Assistant Prof @cmist_cmu

Joined April 2018
Photo by Hannah Bailey (@Hannah_LSBailey)Photo by Hannah Bailey (@Hannah_LSBailey)Photo by Hannah Bailey (@Hannah_LSBailey)Photo by Hannah Bailey (@Hannah_LSBailey)Photo by Hannah Bailey (@Hannah_LSBailey)Photo by Hannah Bailey (@Hannah_LSBailey)Photo by Hannah Bailey (@Hannah_LSBailey)Photo by Hannah Bailey (@Hannah_LSBailey)
I am delighted to announce that in January 2025 I will be joining @cmist_cmu at Carnegie Mellon University as a Tenure-Track Assistant Professor in International Relations and Emerging Technologies! (1/3)
12
2
6
87
18,609
Hannah Bailey retweeted
“The language of unprecedentedness has made ordinary regulation seem inapplicable at precisely the moment when existing law is most needed, @SBarrington_ (@BerkeleyISchool) and @Hannah_LSBailey (@cmist_cmu) write in the Journal of Online Trust and Safety. tsjournal.org/index.php/jots…
1
3
111
Hannah Bailey retweeted
1/ Article now officially out in @ISQ_Jrnl! Every day, we hear about how disruptive technologies are "breaking" international security. AI. Quantum. Biotech. Hypersonics. If alarm bells are sounding, isn't it time to regulate? Yet policymakers almost always wait. 🧵
2
12
1
26
1,575
Hannah Bailey retweeted
When it comes to regulating AI, how can we move beyond the current policy paralysis? In the Journal of Online Trust & Safety, @SBarrington_ (@BerkeleyISchool) and @Hannah_LSBailey (CMIST) propose decomposing AI into three layers. tsjournal.org/index.php/jots…
2
3
150
Ph.D. student @SBarrington_ and @cmist_cmu Prof. Hannah Bailey explore how history repeats itself when regulators confront frontier technologies like AI. Their essay has been published in the Journal of Online Trust and Safety @StanfordTIP tsjournal.org/index.php/jots…
4
1
6
897
Hannah Bailey retweeted
go read @SBarrington_ and @Hannah_LSBailey's excellent piece
Ph.D. student @SBarrington_ and @cmist_cmu Prof. Hannah Bailey explore how history repeats itself when regulators confront frontier technologies like AI. Their essay has been published in the Journal of Online Trust and Safety @StanfordTIP tsjournal.org/index.php/jots…
1
6
562
New in the Journal of Online Trust & Safety: “Regulating the Unregulatable,” with @SBarrington_. AI isn’t as “unprecedented” as claimed. Our three-layer framework helps policymakers spot which parts existing law already covers. tsjournal.org/index.php/jots…
3
112
Hannah Bailey retweeted
Training the model is only half of the work. Excited for Skild to continue scaling up deployments
We just hit 100M ARR within 10 months of starting deployments. We are in factory lines. On construction sites. In kitchens. In data centers. Cleaning. Welding. Building. Cooking. Deploying.
2
1
24
905
Hannah Bailey retweeted
We just hit 100M ARR within 10 months of starting deployments. We are in factory lines. On construction sites. In kitchens. In data centers. Cleaning. Welding. Building. Cooking. Deploying.
122
292
163
2,109
1,689,754
China can build AI, but it can’t install trust eastasiaforum.org/2026/08/05… @cmist_cmu's @Hannah_LSBailey hones in on a fundamental tension between China's ability to build generative AI models and the Party-state's need to constrain information. #China #AI #information #trust
1
2
410
Show it a single human demonstration of a long-horizon task and the robot performs it right away. No fine-tuning, no retraining. Skild has released S1, a robot foundation model where you specify the task with a video demonstration instead of a language instruction. Why video, not language Language works for "hand me the mug." It falls apart for anything delicate or long-horizon. You don't learn to fold a fitted sheet from a sentence. So S1 takes a video demo into its context window and translates it to its own body and scene. What it does - Tasks up to 10 minutes, dozens of sequential steps - Unseen tasks: potting plants, pour-over coffee, kit assembly, pancakes. Skild says even the underlying primitives, like flipping a pancake, weren't in pre-training - Recovers from its own mistakes without being trained to - Improves on flawed demos. Demonstrator drops an egg, S1 doesn't make that mistake. It reads the demo as a goal, not a trajectory to copy.
10
24
2
184
19,096
Hannah Bailey retweeted
Strong scaling laws for one-shot learning in long-horizon manipulation tasks. It's only a matter of time until these models are ready for widespread deployment with very little task-specific data.
Introducing S1, our new foundation model that learns from one example. It can be taught 10-minute long tasks that it has never seen before, from one video prompt without any fine-tuning. Watch S1 operate in real-time via in-context learning:
3
30
1,401
Incredible work by the Skild team (especially my husband @MarcRigter!) - foundation robotics models have arrived!
Introducing S1, our new foundation model that learns from one example. It can be taught 10-minute long tasks that it has never seen before, from one video prompt without any fine-tuning. Watch S1 operate in real-time via in-context learning:
1
12
840
Hannah Bailey retweeted
Introducing S1, our new foundation model that learns from one example. It can be taught 10-minute long tasks that it has never seen before, from one video prompt without any fine-tuning. Watch S1 operate in real-time via in-context learning:
452
1,004
547
6,732
3,563,896
China can build world-class AI. But can it make that AI trusted? In my new piece for East Asia Forum, I argue that Beijing’s commitment to information control imposes a “stability tax” on its global AI ambitions. eastasiaforum.org/2026/08/05…
1
1
305
Todd Hall and I argue that China is acting like a global opposition party, using the weaknesses of US power to campaign for greater international authority. Grateful for this write-up of our article
1
1
2
344
Hannah Bailey retweeted
"Beijing is adopting the tactics of a global opposition party, argues a new paper by Hannah Bailey of the Carnegie Mellon Institute for Strategy & Technology and Todd Hall, director of the China Centre at the University of Oxford." bloomberg.com/opinion/articl…
2
3
316
Beijing resembles a domestic opposition party in advocating its own programme for changing existing institutions. @Hannah_LSBailey and Todd Hall on China’s global opposition campaign: ➡️ go.iiss.org/402G4kt
2
5
213
Thank you for having us!
On this #SOAS China Institute podcast Todd Hall, director of @ox_chinacentre and Hannah Bailey @Hannah_LSBailey, assistant professor at @cmist_cmu join us to discuss China’s global ambitions. Listen here: soas-china-institute.buzzspr…
1
2
938
Hannah Bailey retweeted
1/ Technology is always advancing. So what does it take to future-proof the law? My latest research in @IntOrgJournal argues that efforts to maximize legal clarity can actually undermine the law’s resilience. cup.org/4bX6mKJ
9
17
3
56
7,445