The annual college entrance examination is coming again. Nowadays, gaokao is not only a big event for high school students, but also a hot topic in the science and technology circle.

After last year’s artificial intelligence robot ai-Maths challenged students to take the national college entrance exam, the gaokao, for the first time, ARTIFICIAL intelligence has once again entered the exam with many students — this time to write essays.

The robot will challenge zhang Yiyi, a young writer from mainland China, in the Chinese section of this year’s college entrance examination.

Artificial intelligence VS famous writers

Who will be the winner of this “man-machine war”? The reason that AI can beat Lee Sedol in go and solve math problems in the college entrance examination is based on powerful algorithms and data, but this does not mean that AI has excellent language thinking and emotional thinking and can write touching articles.

Some college entrance education experts believe that computer programmers can analyze data from years of high-scoring essays and “train” robots to produce essays that best meet the tastes of examiners. While robots are not yet able to sincerely express the complex emotions of humans, the time required for grading gaokao essays is short, and even-handed essays can earn good marks, if not high marks.

Although Zhang’s compositions have become popular online over the years, he only got 29 points in his 2014 gaokao essay. The news of “failing writer’s composition” once triggered a continuous hot discussion among netizens. It can be seen that writers who do not follow the fixed rules of college entrance examination composition may not get high marks.

On the other hand, some insiders believe that the core of machine writing is to analyze information with big data, rearrange and combine it with fixed algorithms, and then present it in a specific format, which results in fixed writing mode and narrow application scope. Writing robots beat humans in “numbers” and “efficiency,” but they lack flexibility and creativity. After all, a machine is still a machine, and it doesn’t have the same capacity to think as a person.

How does artificial intelligence write articles

How does artificial intelligence write articles? There are three types of writing robots on the market, all of which rely on “natural language processing” technology.

The first type is template writing, that is, using an article template to fill in the structured information in the database, including specific numbers, percentages, etc., for the traditional sense of “scripted” work.

The second type is post-extraction collation, which is to analyze the long natural language text containing useless information, extract the key information in the article, reorganize the language and output it.

The third method is self-generation. The usual method is to train the language model, conduct mathematical modeling of the language, and then use its own way to generate the words that the algorithm thinks “can best reflect the meaning of the original text and look like a sentence” without using the original words of the original text.

Application of robot writing at present most is news dissemination domain, the robot can help reporters or editors to assume part of the news work, such as sports scores progress, athletes, or financial news of the stock market changes, and so on, this information manually processing is very cost time and energy, belongs to the work of labor intensive, And the staff only need to customize the program and delivery template, the application of robots to deal with, is a good way.

However, judging from the current progress of technological development, the application of AI robots for novel, poetry and other complex literary style creation is not very smooth. Most of the sentences written by AI programs are not smooth. While the industry has claimed that AI can produce universal works, human intervention is needed to improve.

To sum up, no matter whether artificial intelligence can write a colorful article or what the result of man-machine war is, the essence of writing is to return to the essence of writing. A touching work depends not on the gorgeous rhetoric piled up under the machine writing set, but on the sharp observation and innovation of people under the independent thinking. AI write college entrance examination composition in today can hit what kind of score? We’ll see.

Above, Chen Ruchu’s humble opinion! I’m sorry if I offended you.

For more information on artificial intelligence, please visit Giiso Zhisou :www.giiso.com. Thank you