Artificial intelligence in the traditional statistical sense is largely

In the statistical sense, traditional artificial intelligence controls samples to a great extent by statistical methods.

Statistics is a discipline that discovers laws from data and plays a vital role in the development of artificial intelligence. It can be simply understood that artificial intelligence is a statistical application, and the decisions made by our current artificial intelligence are all "experiences" obtained through a large amount of data analysis.

In perceptual technology, AI has achieved 95% and 99% accuracy in speech recognition and image recognition respectively. Whether it is statistical language model or deep network model, it mainly relies on mathematical statistics methods such as probability theory, Bayesian theory and optimization theory as basic tools.

The essence of strong artificial intelligence stage is data-based learning and reasoning, and statistics is the "wheel" to push it forward. Statistics is not only the basis of solving algorithm and model problems, but also the statistical thinking has great influence on the data acquisition, algorithm development and result analysis of the whole man-machine cooperation.

Although artificial intelligence is inseparable from statistics, statistics are not the whole of artificial intelligence.

If we want to build a robot that can play the violin, we need not only statisticians and artificial intelligence experts, but also mechanical engineering experts, musicians, behaviorists and other talents. Let's work together. In today's world, cross-border and integration are king, and statistics are the glue that binds them together.

As enterprises collect more and more data from expanding sources, in order to analyze and interpret these data, enterprises and organizations will need to hire more people (statisticians) specially trained in these analysis. In the past, people used the information gap to earn excess profits, and the future excess profits will be created by super data analysis-statistics.