Pinned post
Delighted to share that my latest article with @bethanygoldblum in @JPR_journal "Evidence of the unthinkable: Experimental #wargaming at the #nuclear threshold" is now live: journals.sagepub.com/doi/10.1177/00… 1/9
Andrew Reddie
@areddie89
Faculty @GoldmanSchool @UCBerkeley; Founder @BerkeleyRisk; Dir @BtGProjectDC; Affil @UCIGCC; Term Member @CFR_org; Fmr @Politics_Oxford.
3.3K Following 2.2K Followers 4.3K Posts
Latest posts 5
From #CRISPR to #AI, I spend much of my time thinking and writing about technology’s impact to the right of its development. Here, Jon and I focus on the forces driving #innovation to the left.
Here's to the R1s and FFRDCs that underpin our venture economy...
Here's to the R1s and FFRDCs that underpin our venture economy...
It’s not just about startups. The secret to America’s tech edge starts upstream. ow.ly/Y3m630sNExSView quoted post on X
How the government responds to the competition between the US and #China will be one of the most important challenges for the new administration. We need to get it right...
lawfaremedia.org/article/the-ha…
www.lawfaremedia.org The Hawley Act Threatens AI Innovation Senator Hawley’s new bill seeks to cut U.S. AI ties with China but risks stifling innovation and hurting U.S. technical dominance instead. lawfaremedia.org/article/the-ha…
Ritwik Gupta and @areddie89 explain what is in Senator Hawley's new AI bill aimed to stop collaboration of AI tech between the U.S. and China and why it will stifle American AI progress and innovation.View quoted post on X
Delighted to see the latest from @BerkeleyRisk w/ @Ritwik_G and @LPWalkr in the wild. Using the recent release of Tencent's Hunyuan-Large model, the paper begs the question as to whether hardware-oriented #AI #export controls are doing what policymakers want...
China is bypassing U.S. export controls on GPUs, but how?
In my latest paper, "Whack-a-Chip," we provide the first public evidence of how China is using advances in ML to make old chips useful for SOTA results.
👉🏽 arxiv.org/abs/2411.14425
@areddie89 @berkeley_ai @BerkeleyRiskView quoted post on X
As anyone who has been following various #AI #governance debates knows, a great deal of ink has been spilled on whether and how to think about limiting the size of these models.
But is this the right approach? In our latest, we suggest not.
arxiv.org/abs/2409.17216
But is this the right approach? In our latest, we suggest not.
arxiv.org/abs/2409.17216
arxiv.org Data-Centric AI Governance: Addressing the Limitations of... Current regulations on powerful AI capabilities are narrowly focused on "foundation" or "frontier" models. However, these terms are vague and inconsistently defined, leading to an unstable...