Foolish Exclusions
China and the Bletchley Declaration on AI
At the conclusion of the Second World War, debates raged on how best to regulate the destructive power of the atom. Splitting it had been used most savagely against the populations of Hiroshima and Nagasaki in August 1945, thereby ending, to date, the globe’s costliest war. Visions also abounded on the promise and glory of harnessing such energy. But the competitive element of pursuing nuclear power never abated, and attempts at international regulation were always going to be subordinate to Realpolitik. Yet even at such a tense juncture in human relations, it would have been absurd, for instance, to have excluded such a major power as the Soviet Union from such discussions.
Over the first few days of November, at Bletchley Park in Buckinghamshire, we saw something akin to that parochial silliness take place regarding discussions on the safe development of artificial intelligence (AI). While the People’s Republic of China was not entirely barred from attending proceedings at UK Prime Minister Rishi Sunak’s widely advertised AI Safety Summit, it was given a shrunken role.
The very fact that China has any role to play was enough to send Liz Truss, Britain’s stupendously disastrous, short-lived former Prime Minister, into a state of spluttering agitation. In a failed effort to badger her successor via letter to rescind the initial invitation to Beijing, she revealed how “deeply disturbed” she was that representatives from the evil Oriental Empire would be participating. “The regime in Beijing has a fundamentally different attitude to the West about AI, seeing it as a means of state control and a tool for national security.”
Seeing the Middle Kingdom was uniquely disposed to technological manipulation – because liberal democratic governments apparently have no interest in using AI for reasons of controlling their subjects – she failed to see how any “reasonable person” could expect “China to abide by anything agreed at this kind of summit given their cavalier attitude to international law.”
Sunak, to his credit, showed some mettle in parrying such suggestions. In a speech delivered on 26 October, he owned up to his belief that China needed to be invited. “I know there are some who will say they should have been excluded. But there can be no serious strategy for AI without at least trying to engage all of the world’s leading AI powers.”
Despite this, Sunak was hardly going to give Beijing unfettered access to each and every event. Some minor form of segregation would still be maintained. As UK Deputy Prime Minister Oliver Dowden stated with strained hospitality, “There are some sessions where we have like-minded countries working together, so it might not be appropriate for China to join.” Largely because of that sentiment, Chinese delegates were, for the most part, excluded at public events for the second day of the summit.
From within the summit itself, it was clear that limiting Beijing’s AI role would do little to advance the argument on the development of such technologies. A number of Chinese delegates attending the summit had already endorsed a statement showing even greater concern for the “existential risk” posed by AI than either the Bletchley statement or President Joe Biden’s executive order on AI issued at the end of October. According to the Financial Times, the group, distinguished by such figures as the computer scientist Andrew Yao, are calling for the establishment of “an international regulatory body, the mandatory registration and auditing of advanced AI systems, the inclusion of instant ‘shutdown’ procedures and for developers to spend 30 per cent of their research budget on AI safety.”
For the Sinophobe lobby, one awkward fact presents itself: China has made giddy strides in the field, having made it a policy priority in its New Generation AI Development Plan in 2017. The policy goes so far as to acknowledge, in many ways providing a foretaste of the Bletchley deliberations, the need to “[s]trengthen research on legal, ethical, and social issues related to AI, and establish laws, regulations and ethical frameworks to ensure the healthy development of AI.” Some of this is bound to be aspirational in the way that other documents of this sort are, but there is at least some acknowledgment of the issue.
Precisely for its progress in the field, China is being punished by that other contender for AI supremacy, the United States. Despite some forced sense of bonhomie among the delegates, such fault lines were nigh impossible to paper over. On 17 October, the US Department of Commerce announced that further restrictions would be placed on advanced AI chips along with the imposition of additional licensing requirements for shipments to 40 countries to prevent resales to China. One company, Nvidia, was told directly by the department that it had to immediately cease shipping A800 and H800 chips to the Chinese market without licensed authorisation from the US.
The final Bletchley Declaration opens with the view that AI “presents enormous global opportunities: it has the potential to transform and enhance human wellbeing, and prosperity.” With that in mind, the signatories affirmed “that, for the good of all, AI should be designed, developed, deployed, and used, in a manner that is safe, in such a way as to be human-centric, trustworthy and responsible.” But the vision risks being irreparably fractured, contaminated by such fears so crudely expressed by Truss. The view from the signatories present is that the AI frontier presents ecstatic opportunity and potential calamity. But how that vision is duly realised will depend on what is decided upon and whether those rules will be observed.