How the Trump Presidency could boost the prospects of Safe and Democratic Global Governance of AI
After Trump's election, we live in a whole new world in respect to the future of AI and its governance.
Trump, together with Elon Musk, has accrued immense power over the future of AI safety and control. The US “Deep State”, Jinping, and Putin are also going to be important decision makers, alongside companies Microsoft, Blackrock Meta, and a few powerful states, but much less than those two.
While Donald Trump's unpredictability, instinctive attitude, and strongman style are reasons for concern for AI governance, Trump's foreseen non-ideological and practical approach to foreign policy could represent a significant improvement over his predecessor, who engaged in an ideological winner-take-all crusading AI race against China. Elon Musk has been ambivalent, very aware and vocal of the risks, but also often declaring there is no way to control it anymore.
Given the continued raise in AI capabilities and investments, we can hope that an increase in the awareness of the immensity and urgency of the safety risks, driven by US security agencies, will lead Trump - along with Jinping and Putin - to decide to create a global treaty-organization strong enough to manage AI safety risks.
Given the fast-rising awareness of the difficulties of enforcing AI non-proliferation without wide global consent, will lead them to pursue an effective tracked treaty-making process, which is at least somewhat a globally democratic treaty-making process, and resulting in a global governance of AI.
We were there once before with nuclear technologies and we almost succeeded. In 1945-6, top atomic scientists and security agencies lead the presidents of the US and Russia to propose the bold Baruch Plan and Gromyko Plans, to bring all dangerous AI research, arsenals, assets, facilities, and supply chains under exclusive international control, without any state's veto.
While admittedly a moonshot, perhaps the most effective contribution we can had as NGOs is to:
(1) Lobby relevant public officials and agencies, especially top security and intelligence advisors, to convince their Presidents of the the ineluctability to pursue with the utmost urgency and care a “Baruch Plan for AI” , as they did in 1946; and
(2) Promote the enactment of much earlier, broader, and deeper treaty-making (military, scientific and political) by adopting an effective and timely model process, enabling them to find a strong binding agreement among many states in due time, and avoid the failure of the Baruch Plan.
