Recent data shows that ChatGPT traffic has dropped by 10% in the past 6 months. This raises the question of whether the development of large models has reached a bottleneck. We analyzed this and came to the following observations.
First, ChatGPT has some problems in terms of practicality. The text content it generates is often too templated, and the homogeneity of sentence patterns and structures is too high. This makes the generated content lack uniqueness and emotion, and it is difficult to arouse the interest of readers. Compared with the boring content commonly found in university textbooks, the articles created by ChatGPT are more boring and lack the unique style of the author. Therefore, the content generated by ChatGPT is often difficult to attract the attention of readers, which is a major problem in its practicality.
Second, ChatGPT is not very accurate in processing in-depth professional information. For some more complex and professional content, ChatGPT's answer may not be accurate enough, and even further verification and verification are required. Although ChatGPT can provide users with some ideas and routine practices, in practical applications, additional learning and data review are still required. Especially in work such as writing code, the code generated by direct copying often has a high error rate. If you cannot review the details of the code and understand its principles, errors will be more difficult to troubleshoot.
Third, the use of ChatGPT is relatively complicated and expensive. The actual frequency of use is relatively low, and its use value is limited for people who work on writing materials, and it is not an essential tool for technical staff. Judging from the actual effect of improving work efficiency, its price is high and its frequency of use is limited. For now, ChatGPT is still equivalent to an intelligent search engine, and the accuracy of the search results cannot be guaranteed.
In addition, ChatGPT has formed a certain blow to some unscrupulous merchants who abuse paper copywriting and other behaviors. This has been welcomed and supported by the majority of users.
Overall, although ChatGPT has achieved some success in some fields, there are still some problems and limitations in practical applications. The development of large models may face certain bottlenecks, and further technical improvements and innovations are needed to enhance their practicability and accuracy.
It just so happened that I made an embedded learning material, which is only given to novices. After all, it contains the learning knowledge and directions of each period of beginners' entry, including programming teaching, video explanations of problems, 800 sets of complete sets and various software installations. package, it is not at the same level as the scattered information on the Internet, knock 6, and send it to you for free