Meet the Politician the AI Industry Is Trying to Stop
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
In this episode of Odd Lots, hosts Joe Wiesenthal and Tracy Alloway interview Alex Bores, a New York State Assembly member who has become the target of a $100 million AI industry super PAC. (08:20) Bores, who previously worked as a data scientist at Palantir, discusses his RAISE Act—legislation requiring AI companies to have public safety plans and disclose critical incidents. The conversation explores the intersection of AI regulation, politics, and technology policy, touching on everything from deepfakes to data centers. (30:00) The episode highlights how AI has already become a major political issue affecting labor markets, electricity costs, water consumption, and wealth inequality.
Speakers
Joe Wiesenthal
Co-host of Bloomberg's Odd Lots podcast and editor at Bloomberg. He brings extensive financial journalism experience and frequently covers technology, markets, and economic policy issues.
Tracy Alloway
Co-host of Bloomberg's Odd Lots podcast and senior reporter at Bloomberg. She specializes in financial markets, technology, and policy analysis with a focus on how these intersect with everyday economic realities.
Alex Bores
New York State Assembly member and candidate for the 12th Congressional District representing Manhattan. He has a master's degree in computer science and previously worked as a data scientist at Palantir from 2014-2019, eventually becoming one of five overall leads of the government business. (45:00) He is the first Democrat elected in New York State with a degree in computer science and has passed 27 bills during his three years in office.
Key Takeaways
AI Regulation Requires Technical Expertise
Bores demonstrates that effective AI regulation requires lawmakers who understand the technology deeply. (08:50) His RAISE Act targets only companies spending $100 million on compute for final training runs and models with 10^26 computational operations, showing how technical precision can create focused regulation that doesn't burden smaller players. This approach counters the industry argument that any regulation stifles innovation by proving that knowledgeable lawmakers can craft targeted rules that address safety without hampering development.
Government Must Track Policy Implementation, Not Just Pass Bills
Drawing from his Palantir experience, Bores emphasizes that "the work isn't done when the bill is signed." (35:14) He shared examples like his telemarketing fine bill, which increased actual fines by 4x after raising statutory maximums, and his failed moped registration bill that showed registration numbers dropped from 1,700 to 1,400 because people weren't re-registering. This data-driven approach to governance demonstrates how politicians can use metrics to assess and improve policy effectiveness over time.
Technology Solutions Exist for Deepfake Problems
Rather than expecting humans to detect AI-generated content, Bores advocates for cryptographic solutions like C2PA (Content Credential Provenance Authority). (28:49) This metadata standard can cryptographically prove whether content came from a real device or was AI-generated, similar to how HTTPS solved internet banking security. The key is making this the default option so that content without cryptographic proof becomes suspect, creating a technological rather than human-dependent solution to authenticity verification.
AI Industry Self-Regulation Is Insufficient Without Legal Requirements
While AI companies have made voluntary commitments at White House summits and international gatherings, Bores argues these need legal backing. (16:00) The RAISE Act essentially codifies what companies already claim to do, preventing them from cutting safety measures during quarterly pressure or fundraising rounds. The industry's own estimate suggests compliance would require just one additional full-time employee for companies like Google or Meta, undermining claims that regulation would be burdensome.
State-Level Innovation in Tech Policy Is Critical During Federal Inaction
With federal AI policy uncertain, states are leading on practical tech governance issues. (30:32) Bores highlighted New York's progress on deepfake pornography laws, chatbot disclosure requirements, and click-to-cancel subscription rules. This state-level innovation provides real protection for citizens while federal lawmakers struggle with comprehensive tech policy, demonstrating how local governance can address immediate technological harms affecting daily life.
Statistics & Facts
- The AI super PAC "Leading the Future" initially planned to spend multiple millions against Bores but recently increased their commitment to $10 million. (09:02) This represents a significant escalation in AI industry political spending and demonstrates how seriously they view state-level regulation as a threat.
- Bores has passed 27 bills in his three years in the New York State Assembly, which coincidentally equals the number of bills Congress as a whole passed in 2023. (08:50) This statistic highlights both his legislative productivity and Congress's relative inaction on policy issues.
- Only about 1,700 mopeds were registered in New York State when Bores' registration bill took effect in January, dropping to 1,400 a month later as registrations expired. (36:57) This reveals the massive scale of unregistered vehicles in New York City and the failure of existing registration systems.
