This Former DeepMind Exec Thinks the AI Arms Race Could End in Disaster
Summary
Former Google DeepMind executive Verity Harding argues that the prevailing "AI arms race" metaphor, fueled by nationalistic rhetoric and intense competition between global superpowers like the US and China, risks a disastrous future for AI development. Harding, who briefed leaders from Barack Obama to Emmanuel Macron between 2016 and 2020, observes a shift from international cooperation to rivalry, exacerbated by events like ChatGPT's November 2022 launch and the Ukraine war. She contends that framing AI as a weapon leads to excessive government control, centralized power, and less safe systems, as evidenced by the Trump administration's nationalist AI rhetoric and export controls. Harding advocates for an internationalist approach, suggesting a "middle powers coalition" involving countries like Canada, France, Japan, South Korea, India, and the UK to foster collaboration and prevent smaller nations from becoming mere pawns in a binary superpower struggle.
Key takeaway
For Policy Makers and AI Ethicists weighing global AI governance strategies, you should actively challenge the "AI arms race" narrative. This framing risks centralizing power, reducing safety, and forcing smaller nations into a binary choice between superpowers. Instead, prioritize fostering international cooperation and exploring "middle powers coalitions" to ensure a more distributed, collaborative, and ultimately safer development path for AI, preserving the "cooperation muscle" before it atrophies.
Key insights
The "AI arms race" metaphor, driven by nationalism and competition, actively hinders international cooperation essential for safe and beneficial AI.
Principles
- Language shapes AI policymaking and international engagement.
- Competition and collaboration are not mutually exclusive in AI development.
- Geopolitical context significantly impacts AI's trajectory and policy outcomes.
Method
Form a "middle powers coalition" to enhance influence and scale, fostering collaboration and preventing a binary superpower AI race.
In practice
- Advocate for international AI governance frameworks.
- Explore multi-national partnerships for AI research and development.
- Challenge nationalistic AI rhetoric in policy discussions.
Topics
- AI Geopolitics
- International AI Cooperation
- AI Governance
- Middle Powers Coalition
- Export Controls
- DeepMind
Best for: Policy Maker, AI Ethicist, Director of AI/ML
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by WIRED - Ai.