Chinese AI and computer vision developer SenseTime’s co-founder and CEO, Xu Li, delivered a keynote address at this year’s World Artificial Intelligence Conference (WAIC) in Shanghai on Thursday, underscoring the potential for AI technology to unlock human potential for scientific discovery and innovation.
Specifically, Xu explained how general AI, which mimics human cognition, can perform previously unmanageable tasks by relying on large amounts of computing power instead of fixed volumes of labeled data. This, he said, vastly expands the potential of human intelligence. “Humans have a limited understanding of the unknown world. But in today’s AI era, machines can make predictions, which can help us discover scientific truths earlier and explore them faster. AI has brought about this new innovation paradigm,” Xu said.
The first areas to be significantly augmented by machine conjecture technology include medical research as well as life and material sciences. AI technology must develop in accordance with ethical governance, Xu affirmed, with the ultimate goal of implementing advanced AI to foster social progress.
Playing on the words of famed British mathematician Alan Turing in 1950—Can machines think?—Xu asked the WAIC audience to consider a simpler proposition: “Can machines formulate suppositions?” The answer is unequivocally yes, according to the SenseTime co-founder. General AI, which he describes as “interactive cognition,” will be more potent in a diverse range of applications than the currently popular version of AI technology that relies on sample data. For example, in less frequent use cases like fire emergencies, the quantity of training data may not be sufficient for the optimal application of AI.
“Going forward, I believe that AI and machine prediction will help us make the next scientific breakthroughs comparable to Newton’s laws of motion or Einstein’s revolutionary ideas,” Xu said.
Read this: China’s four tigers of AI are torn between hyped valuations and profitability
KrASIA Connection features translated and adapted content published by 36Kr. This article was originally written by Su Jianxun for 36Kr.