Julia Stoyanovich

1K posts

Julia Stoyanovich

Julia Stoyanovich

@stoyanoj

Comp sci and data sci prof at NYU. Responsible data science, evolving graphs, preferences / voting. Director of the NYU Center for Responsible AI

New York, USA Katılım Nisan 2009
565 Takip Edilen2.1K Takipçiler
Sabitlenmiş Tweet
Julia Stoyanovich retweetledi
Amfetaminolohije
Amfetaminolohije@amfetaminolohie·
Beograđani, predsednik gricka jabuke na Instagramu i gleda kako njegovi huligani u policijskoj opremi biju vaše studente, vašu decu. 🙏
131
745
3.2K
127K
Julia Stoyanovich
Julia Stoyanovich@stoyanoj·
@thehill @nyutandon @AIResponsibly In 2014, I hope that US Congress will 1) mandate disclosure about the use of #AI, starting with the federal govt; 2) mandate a comprehensive inquiry into the environmental impact of AI; 3) substantially increase federal investment in #ResponsibleAI research, education & training
English
1
0
0
460
Julia Stoyanovich
Julia Stoyanovich@stoyanoj·
@NSF @NIH Finally, I recommend that Congress should mandate a comprehensive inquiry into the environmental impact of AI. And it should financially disincentivize unnecessary or unproductive use of AI to limit these impacts. See my full statement for more r-ai.co/AIImpactForum (8/n)
English
0
1
2
360
Julia Stoyanovich
Julia Stoyanovich@stoyanoj·
Second, I recommend that Congress should substantially increase federal investment in #ResponsibleAI research, education and training, and that it should direct that funding to academia through @NSF @NIH and other agencies. We cannot wait for hand-outs from the industry. (7/n)
English
1
1
2
416
Mark D. Levine
Mark D. Levine@MarkLevineNYC·
There is enormous fear in NYC’s Jewish community ahead of potential protests tomorrow. Know that NYPD is taking this extremely seriously. While there are currently no known specific threats, 1000s of officers will be mobilized thru Sunday at Jewish institutions and other sites.
English
39
60
391
245.6K
Julia Stoyanovich retweetledi
Arvind Narayanan
Arvind Narayanan@random_walker·
LLM security is like an onion. Each layer must treat the ones outside it as untrusted. To defend against prompt injection the model developer must reason about the user-attacker security boundary, which they are far removed from. It's like balancing sticks stacked end to end.
Arvind Narayanan tweet media
English
9
86
343
91.3K