00:00Shilat Simawala is an Under-Secretary-General at the United Nations.
00:04It is important to understand that in this era in which we live in,
00:11the issue is not who is leading, it's how do we create mechanisms
00:16so that we bring everybody together so that we can be able to do,
00:23to make access to AI technology possible,
00:28so that we can be able to do capacity building, to make sure that it is broadened,
00:34so that we can be able to democratise AI through international governance.
00:40Yes, we do know that globally China and the United States are leading,
00:45but I think out of this process it is quite clear that for AI to work for the greater good,
00:53all of us must be involved.
00:54The President Xi says AI should not be a solo performance by a single country,
00:59but a symphony of international cooperation.
01:02How close is China to being the conductor of that orchestra?
01:08Well, I mean, I think in many ways China is a conductor of that orchestra.
01:17Many AI models that are affordable are coming out of China, which is good.
01:22The issue of AI efficiency was an issue that was big lost,
01:27especially in the race for more compute power.
01:33We can do much more with less, and China has led the way.
01:39I think the issue of open source, the open models, China is really pioneering that,
01:47and therefore I have to congratulate the government of China for doing that.
01:54But I think it's important as we move forward that we need to be in concert
01:59with what is happening at the United Nations,
02:03what is happening at the regional levels,
02:07whether it is at the African Union, whether it is at the European Union,
02:11whether it is in the Americas,
02:13because collectively we can actually do much more than if we do it on our own.
02:19Now, some Western companies, including Siemens and Airbnb,
02:23are switching to Chinese models.
02:25How much of that shift is about innovation,
02:29and how much is it about the risks of relying too heavily on U.S. technology
02:33when there are export controls to contend with?
02:37I think actually the big driver is cost.
02:40I think the big driver is cost,
02:42and everybody is realizing that if we pursue AI in a way that is not cost-efficient,
02:51then we are on a wrong trajectory.
02:54And of course, research is showing that models that are coming out of China
03:01are cost-effective because of computational efficiency
03:07in the deployment of algorithms,
03:09in the utilization of data,
03:11in the way in which we compute,
03:13and in the way in which applications are being deployed.
03:17And I think this is the right move.
03:20I think it needs to be commended,
03:23but I'm absolutely encouraged by the formation of an organization
03:29that is going to extend those advantages to the rest of the world.
03:34I know that WAIKO,
03:37the organization that came out of this conference,
03:41has already 29 founding members.
03:45The big success is when all 193 countries
03:50actually participate and cooperate with this very important organization.
03:54And we do that as many of the most of the שה
03:54Here we go.
Comments