Artificial Intelligence promises to transform the economy and people's lives across the world and major countries such as China and the US are racing to develop cutting-edge AI systems. But how is China fairing in the field? Yang Chengxi sits down with computer scientist Joseph Sifakis, a Turing Award winner, to get his views.
JOSEPH SIFAKIS Turing Award Winner "The Challenge here is how to integrate AI systems, LLMs for instance, with traditional systems. So I think that for this, China is very well positioned because you have a strong industrial basis. You should develop an AI that is very, very different from the AI we have today. And the purpose will be to make products, services smart, and extending the functionality of existing products."
YANG CHENGXI CGTN Reporter "And what do you think about China's regulatory approaches?"
JOSEPH SIFAKIS Turing Award Winner "Yeah, so China has a different approach from the European here, vertical approach. You have a regulatory frameworks that are quite good, I think. For different types of systems, we have regulations for a recommended system or for a classifier system. So for different types of systems."
YANG CHENGXI CGTN Reporter "Many countries are competing in AI, what do you think are the key factors for the industry's development? And how do you think China is doing so far?"
JOSEPH SIFAKIS Turing Award Winner "I think that China should strive to develop a different kind of AI. I'm working currently for industrial applications, self driving cars, et cetera. We need a different AI. We cannot or LLMs are not enough, or we should make LLMs more robust. How we make them more robust? It's a technical question. But it's clear that AI, as it is today, is not good enough to be used in industrial processes. This should be a priority for China. If you want to preserve your position as an industrial power, you should care about that."