AI and Geopolitics with Angela Merkel’s Nerd-in-Chief, Kirsten Rulf

October 19, 2022·38 min·3 clips

Kirsten Rulf, Angela Merkel's nerd-in-chief, reveals why autonomous cars are still failing due to seagulls and human bullies.

The episode opens with Rulf describing her path from Harvard research on autonomous vehicles to leading digital policy at the German Federal Chancellery under Angela Merkel starting in 2019. Her portfolio spanned AI, quantum computing, digital infrastructure, cybersecurity, and government upskilling. She recounts how autonomous vehicle testing in Boston was repeatedly disrupted by seagulls whose random behavior the AI could never classify, and how humans in San Francisco began deliberately bullying self-driving cars once they realized the vehicles would stop near people. She estimates fully autonomous taxis are a 2030s prospect, and argues autonomous public transit is a more realistic near-term goal. Rulf then attempts a working definition of AI as high-level pattern recognition over large datasets, noting that even the EU has spent over a year arguing about a formal definition for its AI Act. She describes the most advanced machine learning systems as ones that identify their own problems and teach themselves solutions without being told what to solve. The conversation turns to a Google researcher who became so alarmed by his conversational AI's simulation of consciousness and emotion that he left the project and publicly warned against releasing it. Rulf and host Bez-Benkreder discuss GPT-3 and DALL-E 2, noting their capacity to disrupt professions from law to graphic design. Rulf stresses that regulatory timelines, such as the four years to pass GDPR and another five to implement it, are fundamentally mismatched with the pace of AI development. She identifies three priority areas for regulation: global governance frameworks, the political nature of training datasets, and synthetic data that is already outrunning oversight. On the geopolitical map, Rulf places the US ahead in language AI, China ahead in facial recognition, and flags Israel and Russia as significant actors often overlooked in Western discussions. She describes China's extraterritorial data laws as a mechanism to potentially extract data from companies it has invested in globally, calling this one of the most dangerous and underreported dynamics. Russia's AI strengths, she argues, lie at the intersection of cybersecurity and automated warfare, and she cites security expert Bruce Schneier's warnings about AI as a cyber warfare actor that humans cannot out-game. Rulf quotes Putin's statement that whoever leads in AI will rule the world, calling it a genuinely alarming signal. On the optimistic side, she highlights AI's role in pandemic response, precision medicine, cancer research, and climate action, including a Vienna university building designed entirely by AI as the world's first carbon-negative structure. She closes by expressing hope that the AI research community, which she describes as still relatively small and largely ethically conscious, will continue to self-regulate and speak out about risks even at personal cost.

As heard by us

A sober, grounded look at why AI governance lags the technology and the private sector driving it.

Kirsten Rulf brings direct experience from the German Chancellery and earlier work on autonomous driving, which gives the episode unusual institutional credibility when it turns to AI regulation, public power, and the distance between fast-moving technology and slower lawmaking.

Read the full review in PlayNext →

Why you'd press play

If you want AI explained through the limits of real-world government, start here.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from The Big Picture