Stop using fake ChatGPT! True and false ChatGPT, reading this article is enough

Stop using fake ChatGPT, real and fake ChatGPT, just read this article.

In this article, we will discuss the following questions:

1. Is the ChatGPT you are using real or fake? Is this important?
2. Why can’t ChatGPT be pirated?
3. How to distinguish between real and fake ChatGPT?
4. ChatGPT’s official website cannot be opened. What are the domestic alternatives? ?

1. Why can't I use fake ChatGPT?

In March of this year, OpenAI released the 4.0 version of ChatGPT, which marks that the ability of artificial intelligence in a large number of fields has surpassed ordinary people. For example, American lawyers are a highly respected and well-paid profession, and GPT 4.0 outperforms 80% of human lawyers.

The previous artificial intelligence such as Siri and Xiaoai is actually more like artificial mental retardation. ChatGPT4.0 has become the artificial intelligence with the largest knowledge reserves and the highest IQ on the earth. This wave of artificial intelligence breakthroughs has properly transformed artificial intelligence from Forrest Gump with an IQ of 85 to Einstein with an IQ of 160.

So what does this have to do with using real and fake ChatGPT?

This time, we chose to use artificial intelligence because of its high level of intelligence.

If fake ChatGPT is used, on the one hand, there may be a risk of privacy leakage, on the other hand, the level of intelligence may not meet the demand, resulting in delays in work. Think about our original intention of using ChatGPT, we need to interact with highly intelligent artificial intelligence, get the information we want, or help us with our work, so choosing the real ChatGPT becomes very important.

2. Why can’t ChatGPT be pirated?

Someone may ask, we have used pirated versions of Windows and Office for many years, why can't we use pirated versions of ChatGPT?

This requires understanding the nature of this wave of AI breakthroughs.

The essence of this round of artificial intelligence breakthroughs is the miracle of the big language model, which is the difference between stand-alone games and online games.

In the past, there were a lot of pirated software and games in China, which made it impossible for Chinese software and game developers to survive. But Chen Tianqiao of Shanda turned his back on a game called "Legend", because Legend is not a stand-alone game, but an online game that is mainly calculated on the server.

The emergence of ChatGPT, just like the emergence of online games, he does not even have a client, just use a browser to visit the website directly. Much of its calculations are on the server. Its intelligence is realized through computer cluster computing on the server side, and the calculation results are transmitted to your local area and displayed on your browser.

Due to relying on server-side computing cluster computing, small companies with such capabilities simply cannot do it. Therefore, only industry giants have the opportunity for large language models such as GPT. In other words, only top companies such as Google, Microsoft, Amazon, Baidu, Ali, Tencent, Huawei, etc. can do it.

(Microsoft invested 10 billion U.S. dollars in OpenAI, one of the reasons is that it itself is the top three cloud service providers in the world)

Therefore, ChatGPT is only true and false, not pirated.

3. How to distinguish between true and false ChatGPT?

We can’t use piracy and want to communicate with the most powerful artificial intelligence at present. How should we distinguish between true and false ChatGPT?

This requires an understanding of how ChatGPT provides its services. There are three ways in which OpenAI's ChatGPT provides external services.

The first way is to directly visit the official website of OpenAI. As long as you use your computer to open the corresponding webpage, you can get the service of ChatGPT.

The second way is to access through APP. Currently, ChatGPT provides an iOS version of the client.

The third way is through API access. For some third-party websites, such as POE, etc., they also integrate ChatGPT services, and we can access ChatGPT through these websites.

 

The first two types are not accessible in China. Then only the third one remains.

For some conscientious companies, they are connected to the OpenAI service, but it is generally more expensive.

Why is it expensive? Because the interface of ChatGPT is charged.

Every time ChatGPT's interface (API) is called, he has to settle the fee with OpenAI. So every visit, that is, every word spoken has a cost. Adding the server costs and other service costs of these companies, the final services are generally more expensive, ranging from tens of dollars to hundreds of dollars a month. However, after paying for it, I actually used the ChatGPT3.5 version provided by OpenAI for free.

And OpenAI's ChatGPT4.0 version is simply too expensive. It is almost equivalent to a few cents to one yuan to say a word, which is much more expensive than direct access to the official website. Even if you use the back-end service to access ChatGPT's API, it is too expensive. This is still good, we just paid a little more money.

Because of the high cost, some unscrupulous providers did not really use the ChatGPT interface (API), but used some open source software to build the so-called ChatGPT service themselves. These services may look like the real ChatGPT on the surface, but in fact their capabilities are far inferior to the real ChatGPT, and there may be a risk of privacy leakage.

Using these fake services is equivalent to losing money and delaying things.

So, what should we do when the real ChatGPT service cannot be accessed in China?

4. What are the domestic alternatives?

Unfortunately, we are restricted again, ChatGPT does not provide services to Chinese users.

But it doesn't really matter, we have many excellent domestic alternatives to choose from. Compared with the United States, China's artificial intelligence technology is not far behind. These domestically produced alternative products are not only free, but also provided by large companies. We can use them with confidence and don't have to worry about the leakage of personal privacy.

The leading products include: Baidu's "Wen Xin Yi Yan", Alibaba's "Tong Yi Qian Wen", Tsinghua University's "ChatGLM", HKUST Xunfei's "Spark Model" and so on. The use of these products is also very simple, just apply with your mobile phone number on the official website. While it used to be possible to wait months for approval, the good news is that approval has now been accelerated, sometimes to as little as a day or two.

The performance of these domestic AI products is almost the same as that of the free version of ChatGPT-3.5, and Baidu's "Wen Xin Yi Yan" can also query data online in real time, and even upload PDF files for reading and interaction. These functions are very practical of. More importantly, their support for Chinese is relatively good, and the user experience is also very good.

Take Wenxin Yiyan as an example. He supports Chinese very well. We are equivalent to adding an artificial intelligence assistant. Last weekend, I went on a trip to Shanhaiguan. Every time I went to a scenic spot, I asked Wen Xinyiyan to act as a tour guide and introduce the situation to me. What he did well surprised me.

(Baidu's voice recognition is also good, if you use voice input, he will use voice output)

Domestic, safe, and free artificial intelligence assistants, whether it is work, study, leisure travel or other activities, they can provide great help, go to register and apply.

In addition, the use of these domestic AI products is also a way to support the development of national artificial intelligence. Because the more people use it, the more these models can learn from us, and the product performance will get better and better.

Therefore, whether it is considering artificial intelligence to assist us in our work and provide us with convenience, or considering safety, convenience and saving money, or even a little bit bigger, supporting the development of national artificial intelligence, we recommend that everyone try domestic AI products.

The domestic product registration website is as follows

Wen Xin Yi Yan
https://yiyan.baidu.com/

Tongyi Qianwen
https://qianwen.aliyun.com/

ChatGLM
https://chatglm.cn/

Spark large model
https://xinghuo.xfyun.cn/

All of the above are leading companies, and they only need a mobile phone number to apply for the internal test qualification, and all of them are free.

Guess you like

Origin blog.csdn.net/2301_77531618/article/details/132265838