Scaling Laws

Anil Seth, Professor of Cognitive and Computational Neuroscience and Director of the Sussex Centre for Consciousness Science at the University of Sussex, joins Kevin Frazier, director of the AI Innovation and Law Program at the University of Texas School of Law and a Senior Editor at Lawfare, to examine whether artificial intelligence could ever be conscious.Drawing on Seth’s essay, “The Mythology of Conscious AI,” the pair attempt to define consciousness, develop the strongest arguments for and against machine consciousness, and assess what the current evidence can—and cannot—tell us. They also discuss the risks posed by AI systems that are perceived to be conscious and what AI labs, policymakers, and users should do amid profound uncertainty.

Matt Abrams, technologist, investor, and co-founder of $1 Billion for Trust, joins Kevin Frazier, AI Innovation and Law Fellow at the University of Texas School of Law and a Senior Editor at Lawfare, to discuss the collapse of institutional trust and how AI may accelerate — or potentially help reverse — that trend.The conversation explores Matt’s work to rebuild trust in information ecosystems, interpersonal relationships, and major institutions. Along the way, the duo examine the relationship between AI, information integrity, labor disruption, institutional legitimacy, and democratic resilience.

Matthew Schwartz, founder of Virgo, joins Kevin Frazier, the AI Innovation and Law Fellow at the University of Texas School of Law and a Senior Editor at Lawfare, to discuss how AI is beginning to reshape gastroenterology. Virgo’s model is trained on one of the largest collections of endoscopy procedure videos ever assembled, opening new possibilities for earlier diagnosis, personalized treatment of diseases like inflammatory bowel disease, and faster clinical research.The duo also examine how medical imaging data can generate new “digital biomarkers,” what role AI might play in shortening clinical trials, and how regulators and healthcare institutions are grappling with the promise—and risks—of deploying AI in clinical environments.

Daniel Kokotajlo, former OpenAI researcher and executive director of the AI Futures Project, joins Kevin Frazier, Director of the AI Innovation and Law Program at Texas Law and Senior Editor at Lawfare, to detail his policy recommendation–AI 2040: Plan A. It’s a thorough analysis of a policy pathway to delaying superintelligence, which Daniel and his co-authors think is necessary to ensure that the disruptive effects of highly-capable AI systems do not outweigh the benefits. Kevin asks Daniel to explain scenario scrutiny, address feedback from other AI policy stakeholder such as Tom Davidson, and detail what led him to already alter his estimation of how likely it is that policymakers adopt Plan A.

DZ Kalman, fellow at the Berkman Klein Center, research fellow at the Shalom Hartman Institute, senior advisor at Sinai & Synapses, and host of Belief in the Future, joins Kevin Frazier, AI Innovation and Law Fellow at the University of Texas School of Law and a Senior Editor at Lawfare, to explore what religious traditions can teach AI governance.DZ and Kevin discuss religious responses to AI, Jewish technoculture, kosher supervision as a model for AI certification, and the limits of AI use in faith contexts. They also connect these traditions to contemporary debates about alignment, institutional trust, human dignity.

Kent Walker, President of Global Affairs at Google and Alphabet, joins Kevin to discuss Google's proposal for a two-pronged approach to AI governance: a Frontier AI Regulatory Organization, which like FINRA, would operate as an independent, industry-based governance body overseen by a federal agency and the application of existing laws to widely-deployed, less capable systems.The two also discuss the timing behind this proposal and its relevance in light of the recent Hugging Face / OpenAI incident.Follow Kent: @kent_walker

Assemblymember Alex Bores ran for Congress on a platform that in many ways centered on AI policy. He therefore drew considerable support as well as substantial opposition from interest groups across the country. Though Bores lost in the Democratic primary, his experience is worth exploring in detail.Alan and Kevin dive in with some key questions:what did he hear from voters on AI? are everyday New Yorkers worried about benchmarks being saturated?how did the millions of dollars coming into the race change its dynamics?what does this early episode in the political aspects of AI signal for the future of AI governance?what's the proper allocation of AI regulatory authority between states and the federal government?Follow Alex at @alexbores And, as always, get in touch with Kevin via @scaling_laws and @kevintfrazier

Courtney Cox, Associate Professor of Law at Fordham Law School, joins Kevin Frazier, AI Innovation and Law Fellow at the University of Texas School of Law and a Senior Editor at Lawfare, to discuss her forthcoming article, Hardwiring Hercules?.It’s a great read that builds on Dworkin’s idea of a perfect adjudicator, Hercules. Courtney argues that the debate over a “right to a human decision” has missed a central problem: normative uncertainty. Machines may struggle to explain their decisions, but humans often struggle too—especially when judges are uncertain about what they ought to do. At the same time, the common response that people merely have a right to a “better decision,” whether by human or machine, assumes we know what “better” means. Courtney’s paper challenges both sides of the debate and asks what AI governance should look like when regulators, engineers, and judges cannot simply “hardwire Hercules.”