Unless governments take action, the deputy prime minister is cautioning that artificial intelligence could be a destabilising force to the world order.
Oliver Dowden will inform the UN that the speed of progress could possibly exceed government's capacity to ensure it is secure.
The United Kingdom will serve as the venue for an international assembly centered on the regulation of Artificial Intelligence, set to occur in the month of November.
There is concern that, in the absence of regulation, AI could lead to job losses, an increase in false information, and the perpetuation of prejudice.
Mr Dowden will tell the United Nations general assembly in New York that the starting pistol has been fired to begin a rivalry on a worldwide level in which firms and countries alike will make an effort to stretch the limits as far and quickly as they can.
Currently, regulations around the world are lagging behind the current advancements.
Historically, governments have responded to emerging technologies with regulations- however, in the case of AI, the regulations must be created in tandem with the technology itself.
AI companies should not be allowed to assess their own performance, similar to how governments and citizens should have faith that risks are effectively handled.
Nations must take action so that the public will be assured that the most important national-security worries have been resolved.
Mr Dowden will caution, however, against becoming "stuck in discussions about whether it is a device for good or a device for bad - it will be a tool for both".
Experts have expressed their astonishment at the rapid growth in the abilities of certain AI systems. Prof Andrew Rogoyski of the University of Surrey stated to BBC News, “We’ve seen timeframes significantly shortened.”
Marc Warner, CEO of Faculty.ai, emphasized the necessity to make a distinction between narrow AI (e.g. searching for indications of cancer in medical scans) and general AI.
He mentioned that these algorithms have strong capabilities and, currently, they cannot be anticipated with precision when they will be produced.
Although I am not overly concerned about the current form of technology, I believe it is prudent for government to plan for the more advanced versions that are sure to come, and take action accordingly. It was like I was some kind of crazy person for wanting to talk about this."
"I have been researching AI safety for a decade or so now - and it was only about two to three years ago that people started to take an interest in discussing it. Previously, I was thought to be eccentric for bringing up the topic."
For me, just initiating a global discussion on AI safety is an accomplishment in itself.
At a private gathering of technology giants, in Washington, Elon Musk reported that a consensus had been reached that regulation was required.
Yasmin Afina of the Chatham House international-affairs think tank voiced her opinion that attaining an expedient worldwide settlement would be complicated.
In contrast to the many years of negotiations required to come to a consensus on nuclear weapons, she noted that AI is so intricate and divergent of a technology that it will be challenging to reach an agreement which all can accept.
Ms Afina noted that it is imperative for smaller countries, marginalised communities and members of ethnic minorities to have an influential role in the conversation. "If they are not given the opportunity to participate and express their opinions, they won't even be considered," she added.
The Prime Minister is eager for Britain to be at the forefront; however, the Commons Science, Innovation and Technology Committee cautioned that without swift enactment of a domestic law, the EU's AI Act might be accepted as the global norm, eclipsing any initiatives from the UK.
Warner, who was formerly a part of the AI council which offered advice to government, commented that the UK had the potential to become a major player in the development of technology to ensure AI safety, if sufficient investment were provided.
He noted that the proposed route was a "very practical middle path" seeing as there wasn't a significant amount of funding going into it currently.
Comments