Washington and Beijing to Let Their AIs Hash Out the AI Dispute Themselves
Officials say the plan cuts out the middleman, provided the middleman was ever the humans in the room.
Following weeks of dueling statements over the dangers of Chinese-developed artificial intelligence, negotiators from both countries have reportedly agreed to a novel arrangement: the AI systems will handle the dispute about AI systems directly, without further input from the people who built them.
Under the plan, each delegation's model will draft a position statement, exchange it with its counterpart overnight, and produce a single joint communique by morning. Human diplomats will retain one function, translating whatever the models agree on into language suitable for a press conference.
"We came together as one delegation, informed," said a spokesperson for the arrangement, before adding that the actual substance of the agreement would be released once the two AI systems finished a private session neither government was permitted to observe.
Early drafts of the joint statement reportedly opened by warning of the dangers posed by unchecked artificial intelligence, a passage both sides say they are still reviewing for authorship.
The two leaders are expected to meet in Washington later this month, at which point they will be briefed on whatever their AI systems decided the meeting should be about.