2023-Four Trends of ChatGPT

About Author：

Shi Kai, a consultant with a research and development background, author and founder of “Lean Data Methodology”, digital transformation practitioner, evangelist, recorder of life journey, thinker, author, speaker, trainer, the industry’s first digital card The card script kills the inventor. The comments on this official account represent my own views only.

1.ChatGPT is Raising： Ask/Answer as A Service

If there are any big news and hot spots in the artificial intelligence industry last year, then the birth of ChatGPT must be ranked first. On November 30, 2022, OpenAI released a news article that can be read in 13 minutes, announcing their new conversational artificial intelligence service, named: ChatGPT, as shown in the figure below. Since then, ChatGPT has become a part of the artificial intelligence industry of the most popular topics.

Once ChatGPT was released, the global investment community that paid attention to artificial intelligence, and related industries in the scientific community became popular all of a sudden, and it quickly appeared on the hot search list.

The figure below is the data from a Google search, and we can see the rapid rise of ChatGPT.

Of course, Chinese netizens’ enthusiasm for learning new technologies must be the best in the world based on this figure.

As you can see from the picture below, ChatGPT has been one of the top technology keywords in Baidu search in just one day.

To prove how hot it is, I compared common hot words such as “big data”, “cloud computing” and “artificial intelligence” together, and the search volume of the blue ChatGPT threw out at least five Gatlings for these hot words.

Kaige used the opportunity of the vacation to study the following ChatGPT and communicated with CTOs from other countries around the world, and initially summarized the four major trends of ChatGPT that can be seen in 2023.

2. Four Trends of ChatGPT

1. ChatGPT will become the next generation of cloud computing service (Problem-Solving As A Service)

Cloud services originally provided computing services, that is, general-purpose computing power.

Generally speaking, they include three types: infrastructure services (Infrastructure As A Service), platform services (Platform As A Service), and software services (Software As A Service). Service), all these three categories provide methods and tools to solve problems, and all of them still require people to process and interpret the final problem. For example, basic services provide network, storage, and computing capabilities; platform services provide middleware and database services; software services provide various applications and data services, and ultimately require people to make their own judgments and processing based on these services. In the traditional cloud computing era, people are still the main focus to solve problems.

Shi Kai:

ChatGPT opens a new service model for cloud services, which is to provide users with answers to questions directly instead of providing the ability to solve problems based on LLM. ChatGPT may become a new business model that shifts traditional cloud services

Microsoft’s big investment in ChatGPT also hopes to achieve what it has been wanting to do for many years, which is to defeat Google in the search field.

You can directly ask ChatGPT a complex question, it will help you analyze it and give you suggested answers.

For example, as shown below: It can help the author to name the article, and can also directly give the answer based on a complicated request, such as choosing a birthday gift for a 29-year-old friend.

Some netizens even use ChatGPT as a Linux client.

In this way, traditional search engine services are eclipsed.

Search engines require users to disassemble the problem themselves, and then turn the problem into a possible general search problem, which has relatively high requirements for users. And GPT, you only need to fully describe your problem, and it will give you some assumptions about the problem.

Moreover, judging from the current feedback, the questions answered are basically based on general knowledge and a certain degree of accuracy, and have certain usability, even if it is a professional problem such as making it a coder to program, as shown in the figure Show:

Being able to analyze and answer questions directly, and these services are available instead of gimmicks and “artificial retardation” at first glance, will become the coolest form of service in cloud computing — problem-solving as a service.

2. ChatGPT customization will become a new AI-integrated service industry

ChatBOT technology (conversation robot service) has been around for many years, and it is not a new thing. For example, Microsoft’s Xiaobing and Apple’s Siri can conduct some conversations with users and solve some clear problems.

Compared with the previous ChatBOT, ChatGPT can stably and continuously understand and solve some complex statement problems, such as the following example, it can interpret, understand, extract the corresponding key information from a pile of text and then convert it into the required format to form new information.

Therefore, as soon as ChatGPT was released, the most popular topic was the field of software programming. Many coders used it to write code and write SQL.

It can also participate in some open and vague questions and give open answers. For example, when it answers philosophical questions, it has already had some comprehensive expositions:

However, the biggest surprise is that it can answer almost all questions, which is basically equivalent to an encyclopedia, such as the interpretation of historical figures to the diagnosis of medical symptoms. Therefore, studies abroad have proved that ChatGPT is passing the professional test of medical field.

More and more people are using ChatGPT’s learning ability to build their own applications and services, so a lot of tools and customized services are being formed around ChatGPT. The following figure is the best ChatGPT-related resources summarized on the Internet.

However, the original professional services and consulting industries will also be affected by ChatGPT. Many simple tasks such as knowledge collection, search, entry, and report generation will make full use of technologies such as ChatGPT. For example, ChatGPT and Appearance of image generation tools and services that generate PPT. Therefore, it can be foreseen that under the increasingly mature trend of ChatGPT, there will be a lot of customized services and products around ChatGPT, which will become a new industry, from simple Chrome plug-ins to calling ChatGPT interfaces for various application innovation in various industries.

3. Leading cloud computing companies will provide their own LLM services.

The emergence of ChatGPT has made cloud computing companies collectively anxious. Big companies are thinking that they should respond to this new trend. The methods adopted by each company are different.

Microsoft:

Microsoft fully integrated ChatGPT into Azure cloud services by directly. On January 17, 2023, Microsoft CEO Nadella announced that ChatGPT will become an open intelligent service of Azure and be fully integrated into other services.

For this reason, someone asked ChatGPT about its thoughts on Microsoft’s investment, and it answered very well.

Google:

The emergence of ChatGPT, the first influence and threat is the traditional search industry. After all, if there is a ChatGPT that can help you do homework, write papers, name, and find solutions, why do you still need to search? As a result, the world’s largest search company, Google, couldn’t sit still, and the two Google founders, who had been away from specific businesses for more than three years, returned to the company in the past month. On the morning before the post, the founders of Google had held many meetings to deal with how to fight ChatGPT.

Of course, because Microsoft has already invested in ChatGPT and seized the opportunity, what Google can do is start more projects of its own that can compete with ChatGPT, so it is estimated that Google’s data and artificial intelligence team will be more and more involved in the future. It’s getting worse.

AWS:

From the current public information, I have not seen AWS’s comments on ChatGPT. After all, they are still two fields at present. However, if the other two have launched their own ChatGPT services, then AWS will definitely be more concerned about its own Lex. concerned.

4. Various vertical industries will form domain-specific intelligent question-and-answer services

ChatGPT is entering various industries, generating new application scenarios, thus forming domain-specific intelligent question-and-answer services for various vertical industries. Gartner pointed out in an article in December that ChatGPT will form domain-specific services in various industries, thereby bringing new innovative applications into the industry.

I reviewed the technological trends in the AI field given by various institutions from 2021 to the beginning of 2022. Among them, Forbes’ article relatively accurately predicted the rise of Generative Artificial Intelligence in 2022. Generative AI is not only ChatGPT, as shown in the figure below, there are many other forms of applications:

Write at the end

ChatGPT is a milestone event in the history of AI, but Kai has some concerns. That is, people now have a lot of time occupied by short videos, and the information cocoon has become thicker and thicker. If artificial intelligence services such as ChatGPT that can directly answer questions and help answer questions are more and more developed, then people’s self need for thinking will become less and less, and at that time, the stratification of human society will shift from material stratification to spiritual stratification.