View from Washington: The great digital game continues

Since leaving Google, Eric Schmidt has taken on a type of role once described to me by a Whitehall mandarin as “valuable pest”. For example, he is currently trying to prod Washington into more coherent action around the economic and national security implications of artificial intelligence (AI).

An earlier column looked at how the final report is coming together from the US National Security Commission on AI (NSCAI) that Schmidt chairs. The i’s should have been dotted and t’s crossed by early March. During the commission’s deliberations, Schmidt has worked on several papers and campaigns with broadly similar themes.

Perhaps the most important has been Is China Beating the U.S. to AI Supremacy?, co-authored with political scientist Graham Allison for the Belfer Centre for Science and International Affairs at the Harvard Kennedy School of Government. Allison has himself been getting plenty of attention recently for his book Destined for War. It considers how China and the US can avoid the so-called Thucydides Trap where competition between a rising great power and an existing one leads to war.

Throughout, Schmidt’s thinking on AI has brought together its technological, economic, and political elements and done so elegantly. It has stressed integration at every level and expressed alarm at the current lack of it. Following on from that, he is one of those urging the US to adopt a more internationalist outlook, one that could give the UK the chance to establish some post-Brexit credentials with the US.

But on that last point, what the US wants and needs need to be balanced by likely and basically willing partners.

It can feel strange to suggest but the technological aspects may be easier. But such is politics, even in a world of systems-of-systems.

The latest draft of the NSCAI paper convincingly argues that to match and exceed China’s ambitions towards using AI as the foundation of ‘intelligentised’ military capabilities, the US and its allies need to first acknowledge how serious China’s programme is and then address four primary tasks:

1. Invest more and heavily in coordinated AI R&D across the armed services and intelligence community.
2. Ensure interoperability in data sharing and AI systems functionality across the services and IC.
3. Develop common standards for the test, evaluation, validation, and verification of AI systems.
4. Apply appropriate export controls and intellectual property protections.

With Britain as a prominent member of both the Five Eyes (seen as the most fertile short-term platform for digital collaboration) and Nato, and given its move to become a member of the less formal Quad alliance in Asia (with the US, India, Japan, and Australia), these seem merely attractive but necessary.

However, the geopolitical aspects look trickier.

Chapter 15 of the draft NSCAI report conflates and expands upon the commission’s international recommendations. One of the four main proposals is that the US launch an International Digital and Democracy Initiative (IDDI).

“IDDI… provides an opportunity for the United States and like-minded allies and partners to counter authoritarian uses of AI, particularly by providing alternatives to digital infrastructure projects that are used for illiberal ends, endanger the social cohesion among and between democracies, and threaten collective security,” the draft says. “As international digital and telecommunications infrastructure investment needs continue to grow and China continues to use digital development to export authoritarianism and expand influence, the United States and its allies and partners must join forces to coordinate a strategy that maximizes the impact of government assistance efforts and also catalyses private sector investment to address shared challenges.”

Motherhood and apple pie, yes?

This is much more problematic than it looks. Already, as the recent Sino-EU investment deal and its immediate outcomes show, much of Europe is wary if not thoroughly resistant to taking moves that could be interpreted as aggression by Beijing (although Moscow is also a factor here).

Even the UK may not be easily convinced should Biden adopt something akin to the IDDI, as seems very likely. While it has specific concerns over Hong Kong, a growing China-sceptic lobby in Parliament and has noted ongoing tensions between China and Australia, Prime Minister Boris Johnson has repeatedly cast himself in the role of “sinophile” - with perhaps greater consistency than much else in his premiership. So, it is not a cosy relationship but still one that the PM seems edgy about making ‘too’ antagonistic (there is probably a hefty dollop of post-Brexit realpolitik in the mix as well).

The UK and other US allies then have another concern. Could the US be relied upon to hold its more globally cooperative position in the long-to-medium term? Joe Biden has assumed the Presidency at what could soon turn into unwelcoming times, with challenges posed by both the public health and economic aspects of the pandemic likely to dominate most of his first term in office, on top of his need to play political poop-scoop in the wake of his erratic predecessor.

And the blistering irony behind all this is that if these challenges are handled poorly or prove insurmountable, they could put Trumpism back into the White House in four years’ time. What price internationalism and promoting democracy then?

So, it may make more sense for now to consider only the technical side of AI alignment at the military and even broader geopolitical level. However, this is where Schmidt’s value as useful pest comes into focus.

Not all the ideas are his. He chairs an NSCAI that has an impressive line-up and, as noted, other collaborators like Graham Allison are at the forefront of their fields. But what Schmidt does seem to bring – evident as a theme informing the projects he attaches himself to – is the broad-based systems-level approach that highlights how interconnected all the elements now are. Should you or even can you break down this kind of strategy and hope to achieve any degree of success?

When the final report is published next month, we will get a clearer idea of how much the Biden administration plans to adopt. That should tell us how it intends to approach and invite allies to contribute. That strategy will need to be very persuasive, perhaps more delicately so than US diplomacy has historically found necessary.