China plans to diversify the application areas of deep learning. Moreover, this system, which allows computers to be trained to perform tasks like those performed by humans, such as speech recognition, image identification or the formulation of predictions, will be driven by open-source platforms. Similarly, the forecast is that, in the Asian giant, deep learning, whose most common exponents are virtual assistants such as Siri, Alexa or Cortana, will be extended by large companies on a large scale.
As the manager of the search engine Baidu, Ma Yanjun, reminds us, innovation will lead to the cost of this technology, one of the most important in the fourth industrial revolution, being reduced, a factor that will favour its adoption. In machine learning, that is, automatic learning, human operators must extract characteristics from the input data; however, the models used by deep learning already include this potential. In Facebook posts, for example, deep learning is used to tag people in posts.
Chinese entrepreneurs are aiming to soon make it possible to pay in a shop simply by the face of the shopper. At the moment, one of the most common uses is to identify faces for security purposes. However, the Eastern power is planning to link this development, also linked to artificial intelligence, with quantum computing - based on cubits, a special combination of ones and zeros - scientific computing - numerical techniques to solve social and engineering problems, etc. - and life sciences.
The Chinese government supports this move, as it has incorporated this framework into the 14th five-year plan, which runs from 2021 to 2025. Baidu's PaddlePaddle, which is considered the first deep learning platform in China, provides software developers with the tools, resources and services needed to adopt deep learning where appropriate. And by all accounts, the uptake has been positive, and progress evident, in activities of all kinds.
In fact, PaddlePaddle has received more than four million contacts, for more than 157,000 companies and institutions, which has made it possible to create some 476,000 artificial intelligence solutions for administrations, healthcare organisations, transport, finance, agriculture, etc. Other national and international corporations, such as Intel, Nvidia or Huawei, have established collaborations of a different nature. Baidu's chief technology officer, Wang Haifeng, says the company will "continue to push forward" in this direction, as Huawei Technologies Co. is already doing with its artificial intelligence system, MindSpore.
Consultants from the global firm IDC underline that PaddlePaddle already ranked first in terms of market share in China among deep learning platforms by the end of 2021, ahead of Google's TensorFlow and Facebook's PyTorch. A number of success stories explain this development. For example, in the wake of the covid-19 pandemic, Linking Med, a Beijing-based medical data company, launched China's first open source artificial intelligence model for analysing pneumonia CT scans, powered by PaddlePaddle.