The News
The US and China are set to hold their first high-level talks focused on artificial intelligence safety in Geneva on Tuesday, as the two superpowers compete to lead the AI race.
US President Joe Biden and Chinese leader Xi Jinping agreed last year to hold talks on AI, with a heavy focus on mitigating risk and ensuring AI is “safe.” But experts and officials don’t expect the meeting to result in pledges for cooperation on AI research or development.
SIGNALS
Don’t expect AI-related deals or cooperation
US officials have emphasized that the meeting won’t be a place for deal-making, and will not end with a joint statement or any “deliverables.” The White House was vague on details of the expected discussions, Breaking Defense wrote, instead disclosing what’s likely not on the agenda: AI’s role in nuclear arsenals and election interference, and specific concerns about Chinese hacking. Washington is more focused on expressing broader concerns that China’s AI development could undermine US national security, while Beijing’s aim is to bolster its reputation as “a responsible power who is willing to play a leadership role in key areas of international governance,” a China-focused fellow at the Carnegie Endowment for International Peace said.
Military use of AI a key focus for both countries
The meeting comes as the US pushes to develop AI-integrated military tech as fast as, if not faster than, China. The US Air Force recently held an exercise with a fighter jet piloted by AI, one of hundreds of AI-focused projects at the Pentagon, the Associated Press reported. “Whether you want to call it a race or not, it certainly is,” the vice chairman of the Joint Chiefs of Staff said. “China’s working on it as hard as we are.” In March, the US hosted a conference with representatives from 60 countries that have signed an agreement on the “responsible” military application of AI. China is not part of that list.
China AI developing at breakneck speed amid fierce competition
Chinese companies are still playing catch-up on developing AI infrastructure like semiconductor chips, but they are rolling out new systems quickly. 01.AI, a startup founded by Chinese tech pioneer Kai-Fu Lee last March, rolled out a new app Monday that Lee said is comparable to the advanced GPT-4 model from OpenAI — which also debuted a new model Monday. “The saying in media that China is way behind is not accurate,” Lee told Bloomberg TV. Influential Chinese tech voices are most concerned about talent development: “US undergraduate education has done a good job of cultivating some basic skills in the computer field, but domestic education needs to be strengthened in this regard,” said the head of research at ByteDance, during a March panel hosted by a local nonprofit.