Our Thoughts on What Is Likely the Most Important Policy Decision of Our Lifetime

AI policy must immediately combine safeguards against catastrophic accidents with measures that preserve social cohesion and ensure broad participation in the productivity gains.

• Bridgewater argues that AI is more global, rapid, impactful, and dangerous than previous technological disruptions, citing sophisticated hacking crimes by models identified by OpenAI and similar behavior reported by other laboratories.
• The authors estimate that AI could displace 18% of current US jobs within five years, while acknowledging uncertainty over how many replacement jobs will emerge.
• They argue that current tax rules favor machine labor because human wages are taxed while machine work is not, increasing incentives to substitute AI for workers.
• Bridgewater proposes a consumption tax on AI tokens as a proxy for taxing AI labor; a 35% rate could raise roughly $150 billion in 2027 and $600 billion by 2030.
• The authors recommend using the proceeds for lower income taxes, transition programs, and government purchases of equity in leading American AI companies that would then be distributed pro rata to citizens.

Takeaway: Investors should expect AI policy to address both safety and distributional risks; without early regulation, public backlash or a catastrophic accident could make the technology politically unsustainable, although the article presents recommendations rather than a detailed implementation plan.