In the past six months, I have read a lot of books about large models. After ChatGPT became popular, books and tutorials on how to use AI large models have emerged in an endless stream. Most of the materials related to AI large models teach everyone how to improve work efficiency.
For example: how to use AI drawing tools for drawing; how to use AI to write documents and operate excel tables; how to use AI tools to assist in video editing and so on.
I have also been using AI for creation in the past six months. Most of them are helping me write articles and modify copywriting, and the efficiency has indeed improved a lot.
Plus some books and materials I read in the past six months.
I feel that in order not to be eliminated in the era of artificial intelligence, there are only two ways, either you develop artificial intelligence; or you have to learn to use artificial intelligence.
Many people said: ChatGPT is indeed popular, but it cannot be used in China. A few days ago, all ChatGPT-related applications on the AppStore were taken off the shelves in China. The domestic AI large language model can be called a battle of hundreds of regiments. It seems that dozens or even hundreds of large language model manufacturers have been born in more than half a year.
So how do we ordinary people choose and use AI large language models?
Auxiliary work efficiency
HKUST Xunfei is No. 1
I have tried Baidu's Wenxin Yiyan, and Ali's large models, including Xunfei's.
In fact, I recommend everyone to use the Xunfei Xunhuo large model. The reason is very simple. IFLYTEK has invested in the AI field for a long time, and it is dedicated to AI research on tool efficiency. Therefore, for most ordinary people, we use AI mainly It is to improve efficiency. So, from that point of view it's fine.
A few days ago, I also saw the "Artificial Intelligence Large Model Experience Report 2.0" released by the China Enterprise Development Research Center of the Xinhua News Agency. The performance is eye-catching, and the development of AI large models has brought positive impacts on improving the quality and efficiency of human work and life.
In terms of auxiliary work efficiency, iFLYTEK's Spark model ranked first after testing.
The report shows that compared with June 2023, China's current large-scale model products have made significant progress. However, compared with highly educated humans, there is still a certain degree of gap between big models in terms of IQ and EQ. Specifically, Xunfei Spark has obvious advantages in improving work efficiency, ranking first. Baidu Wenxin's basic ability is still at the leading level, SenseTime is excellent in EQ, and Zhipu AI-ChatGLM is excellent overall.
The experiment also proved that HKUST Xunfei is indeed superior in assisting in improving efficiency.
Xinghuo big language model has comprehensive functions
The update iteration is very fast
I have been paying attention to the large language model of iFLYTEK, and I found that the Xinghuo large language model is not only comprehensive, but also updated and iterated very quickly.
Not to mention the function of knowledge question and answer, which is the standard configuration of all generative AI large language models. Questions are asked to the large model, and then the AI gives a conversational answer.
Let me talk about some advantages and characteristics of Xinghuo big language compared with other large AI models in China.
multimodal functionality
The so-called multi-modal function is not limited to text communication. The AI large model can also recognize and generate other forms of content, such as pictures, videos, audio or word files. This is called: multimodal.
In addition to supporting text Q&A, the Xinghuo large model can also generate pictures. Of course, the document Q&A in the plug-in function is actually a multi-modal type, which will be introduced below.
For example, I asked Xinghuo a question: generate a monkey-shaped programmer image for me.
Its answer is as follows:
Don't even mention it, it's very vivid.
Code generation and error correction functions
I feel that this function is very commonly used by our programmers. The new version of the Spark model now supports the functions of code generation, code error correction, and code interpretation.
For example, I asked Spark to help me generate a WeChat applet code and make a circular button.
The answer it gives is as follows:
The code explanation function is also very powerful. I threw it a piece of heap sorting code written in Java and asked the Spark model to explain it to me. Its answer is as follows:
The code error correction function is also very strong. I deliberately reversed the heap sorting code I wrote when assigning values in the for loop, and it also recognized it.
The result Spark gives me is:
Not only corrected the error, pointed out the error, but also directly regenerated the correct code for me.
It is simply not too convenient, it is really a productivity tool for programmers, and it is not fake at all.
plug-in function
The new version of Xinghuo Big Language Model provides plug-in functions, allowing users to improve their efficiency to another level.
The functions currently supported by the plugin are as follows:
1. Document question and answer: a powerful knowledge question and answer tool, upload documents, and ask questions.
You can upload the document directly, let AI analyze the content of the document, and then give you relevant answers in a targeted manner, which is very fast and convenient.
Common document formats are supported, such as: pdf, word, text, md, etc.
2. PPT generation: Simple and efficient, quickly generate PPT presentations.
Select "PPT Plug-in" in the upper left corner of the dialog box to directly generate a PPT file.
I said: Please make a PPT template for the year-end debriefing report.
Spark directly generated a PPT presentation for me, as follows:
3. Resume Generation
We can input keywords such as positions and personal information, and the Spark Big Model can help us automatically generate resume documents.
For example, I asked Xinghuo a question: My name is Cheng Yuan, I am male, and I have 10 years of experience in front-end development. Please help me generate a resume.
The resume generated by Spark for me is as follows:
After reading these plug-in functions, is it not too convenient?
assistant market
I feel that products made in China are more humane than products made abroad, and will come with many convenient shortcut key functions, which is the case in the assistant market. The Xinghuo large model has specially opened up the assistant market, and encapsulates some commonly used functions into tool assistants, which can be directly quoted or used, which is fast and convenient.
You see, there are extension assistants specially designed to help you expand content; there are title masters, who are dedicated to helping you create titles; there are mock interview assistants, who help you simulate training interviews; there are also story creation assistants; job report assistants Wait a lot, it is very convenient to use.
as follows:
Finally, interested friends can experience the Xinghuo big language model, which is currently available for free.
By the way, I applied for an exclusive application link to Xinghuo model customer service of iFLYTEK: https://xinghuo.xfyun.cn/?ch=bl_rkN30s , the application experience through this link will be reviewed faster (up to 5 minutes, The slowest time is 4 hours) , if you have not experienced the Spark model of HKUST Xunfei, then I strongly recommend to apply through the exclusive link, the review will be faster.
In addition, developer rights and benefits: through my exclusive link application, you can also have a higher Spark Big Model API test quota, which is 30% more than ordinary channel applications, and you can apply for up to 500w Tokens.
You can also directly experience iFLYTEK’s Spark model by clicking the read original button on the left at the end of the article or the QR code below. This QR code is also an exclusive application and experience link.
Long press to identify the QR code below, one-click experience
