One of my friends has loved art since childhood and spent ten years studying it. He dropped out of school and went to a vocational college to specialize in "drawing movie posters," but still hasn't found employment. He feels like he's constantly chasing something, but is afraid that if he's not careful, he'll be left behind by the relentless tide.

Similar anxieties are playing out on a large scale today, as ChatGPT has arrived.

"Even as a '90s kid growing up with new technologies, seeing the development of artificial intelligence to this stage still makes me nervous and shaky," one internet user said.

ChatGPT is an AI chatbot system. Some users have asked it to write poetry, novels, or small-format text, while others have used it to write code. The results you get depend on how you use it. This is a heavyweight product that was released in November 2022 and attracted over one million users within two weeks. According to foreign media reports, a UBS analyst estimated that ChatGPT reached 100 million monthly users in January 2023, completing in two months what TikTok achieved in about nine months of user growth.

It's not just a product; it's a system that third-party companies can use. Microsoft, which is investing billions of dollars in OpenAI, ChatGPT's parent company, has already applied it to Bing and Edge browsers, announcing on February 8, 2023, that it would integrate the GPT-4 model (the upgraded version of ChatGPT's model) into them.

Its ability to process and create information is amazing, but it has also caused panic among a large group of workers. Over ten professions, including writers, data analysts, customer service representatives, and secretaries, are considered to be the hardest hit by the threat to their jobs.

This article attempts to answer some basic questions about ChatGPT, such as what it is, what its strengths are, what it can and cannot change, and why we should not be afraid of it. Some questions may not have a standard answer, but what is certain is that, as foreign internet users have described ChatGPT's arrival, "it's like waking up to news of the first nuclear explosion; you don't know how to react, but you know the world will never be the same."

What is ChatGPT?

ChatGPT is a large language model that has been trained on vast amounts of text data, allowing it to generate text with impressive language capabilities

The name "ChatGPT" combines "Chat," which refers to chat, and "GPT," which is short for "Generative Pre-trained Transformer." This is a type of natural language processing model that is highly effective due to its large-scale pre-training with massive amounts of language data. There have been several iterations, including GPT-1, GPT-2, and GPT-3, with ChatGPT using the GPT3.5 model.

Compared to other chatbots like AI customer service, Siri, XiaoDu, and XiaoAi, ChatGPT is fundamentally different in several ways. For example, ChatGPT is trained on a vast amount of internet text data, which gives it greater accuracy and a broader range of applications. It is also much larger and has more processing power than other models, which allows it to handle more complex language structures and understand complex grammatical and semantic elements. According to Tsinghua University Professor Huang Minlie, ChatGPT has surpassed the conversational ability of 80-90% of people.

From a technical standpoint, ChatGPT has made several advances. First, it has established the prototype for general artificial intelligence, which can support most language-based tasks. Second, it has lowered the threshold for human-machine interaction, allowing ordinary people to participate in AI-generated content. Finally, its "Prompt" mode has the potential to guide AI thinking using examples rather than requiring retraining, thereby reducing the cost of deploying ChatGPT in various applications.

The GPT model's parameter count, which can be thought of as the amount of language data the model has been trained on, continues to rise, with GPT-1 having 117 million parameters, GPT-2 having 1.5 billion, and GPT-3 having a staggering 175 billion. OpenAI is now working on GPT-4, which is expected to be released in the first quarter of 2023, with the potential to further improve the accuracy and capabilities of ChatGPT. However, there is the possibility that GPT-4 may be the last iteration, as it may have reached the limit of available language resources in human history.

This is a pivotal moment in history. Our understanding of creativity and the act of creation has been challenged by ChatGPT, which has, to a certain extent, shattered our previous beliefs.

Before the Industrial Revolution, workers manipulated raw materials such as steel and wood with tools and their own skills to create commodities. But with the advent of machines, workers lost their jobs as machines took over the process of altering materials.

In a similar fashion, many white-collar jobs center around information processing. Workers receive information, process it using their skills and tools like computers and writing software, and create new information. "This isn't too different from the changes of the industrial era," says Wang Sheng.

A straightforward interpretation is that ChatGPT now performs the task of processing and producing information that was previously done by humans.

This shift from professionally generated content (PGC) to user-generated content (UGC) and then to AI-generated content (AIGC) has significant applications in various fields such as code, images, videos, and games, which can enhance the efficiency of the creative process and reduce the cost of content production.

ChatGPT's power can indeed be intimidating.

Before ChatGPT, there were other buzzworthy concepts like metaverse and Web3. However, what sets ChatGPT apart is that it has a clear application scenario and can be easily applied on a large scale.

OpenAI is a model company rather than an application company, so it provides a fundamental ability that can be integrated into various tools used by different businesses. This ecosystem can help promote the industry, and it is the fastest way to do so.

Furthermore, the current costs are not too much of an issue. According to a report by American venture capital firm Ark Invest, completing a GPT-3 training session in 2020 cost $4.6 million, but this cost is expected to decrease at a rate of 70% per year. The decrease in training costs will support large-scale commercialization.

The report also indicates that by 2030, AI can increase knowledge workers' productivity by more than four times. If it is implemented 100%, AI can increase global labor productivity by about $200 trillion, far surpassing the $32 trillion wage bill of knowledge workers.

However, ChatGPT is not perfect yet.

For instance, OpenAI's CEO Sam Altman tweeted that "relying on it for anything important right now is a mistake," and "we have a lot of work to do on robustness and correctness."

ChatGPT still has a lot of room for improvement, such as understanding diversity and inclusivity, exploring how to help the model understand multiculturalism, gender, age, etc. without bias. It also needs to reduce the risk of reckless answers since its answers are based on a large amount of pre-trained text data, which can result in incorrect or inappropriate responses. It needs to enhance its understanding of the context and improve the interactive experience. In addition, it faces some controversies related to policies and regulations.

The widespread commercialization of ChatGPT may take some time, but it is not as important as the fact that the potential of the technology is already evident. The remaining issues can be resolved with time and effort.

For example, some students have used ChatGPT to write their papers and assignments, which has sparked controversy. To address these concerns, OpenAI has recently released an AI-generated content identifier, which is not perfect but provides a solution.

The age of large-scale ChatGPT application is coming, and it may arrive sooner than anticipated. However, its emergence has both advantages and disadvantages. On the one hand, ChatGPT can greatly improve efficiency. For instance, given a set of keywords, ChatGPT can generate a decent article. Companies like BuzzFeed have already started using ChatGPT to create news content.

According to Gartner, by 2027, chatbots will become the primary customer service channel for about 25% of companies. The possibilities for AI-generated content and the AIGC industry are vast. There are many potential applications for ChatGPT in industries such as customer service, creative content, standardized production, and image generation.

However, on the other hand, the technology may also lead to job losses, particularly for professions that involve customer service, telemarketing, content writing, legal assistance, and data analysis. The bottom line is that the impact of ChatGPT on individual professions is difficult to predict. Therefore, it's better to examine the fundamental logic behind the technology, which is the increase in productivity of content creation.

For instance, while ChatGPT can easily produce basic writing, it cannot replace in-depth content creation, such as long-form articles, at present. However, ChatGPT can still help to streamline the research process and reduce the time needed to collect data. Therefore, those who focus on high-level content creation are less likely to be replaced. In conclusion, individuals in the workforce must evaluate the quality of their output and determine if they are susceptible to being replaced.

Learning to Use ChatGPT

The article discusses how the emergence of ChatGPT, a powerful language model, will bring both positive and negative impacts on various industries. ChatGPT has a clear application and can significantly increase productivity, but it may also replace certain job positions, causing job loss in the short term.

The article emphasizes that while ChatGPT may be a cause for concern, it is essential to understand that it is just a tool, and its impact depends on how people use it. In this sense, it is crucial to learn how to utilize ChatGPT effectively to increase productivity and creativity.

The article provides examples of how ChatGPT can be applied, such as in the film industry, where AI can create concept art and storyboards, or in gaming, where NPCs can have different responses depending on the player's actions. Furthermore, ChatGPT can streamline the process of recording meeting minutes and make them easily accessible in the future.

The article acknowledges that ChatGPT may replace some job positions, but it can also create new opportunities in AI-related fields. Therefore, it is essential to recognize the potential of ChatGPT and to learn how to use it to maximize productivity and creativity.

The emergence of ChatGPT has the potential to revolutionize the way we work, but also raises concerns about the future of certain jobs. However, it's important to note that while some jobs may become obsolete, the development of new technologies will also create new jobs and opportunities.

To keep up with the wave of technological advancements, it's important to understand and learn how to use new tools like ChatGPT. Just as email revolutionized communication and computers changed data processing, ChatGPT has the potential to make content creation and other tasks more efficient.

By embracing new technologies and learning how to use them effectively, people can increase their productivity and creativity. In the long term, this could free people from mundane labor and enable them to engage in more fulfilling work.

While it's natural to feel anxious about being left behind in the face of rapid change, the best approach is to embrace new technologies and seek ways to make them work for us. By doing so, we can take advantage of the new opportunities that emerge and stay ahead of the curve.

ILLA Cloud

ILLA Cloud is a low-code development platform with dozens of front-end components and database API integrations. You can use ILLA Cloud to build the front-end interface by dragging and dropping components and connecting to your database or API to complete full-stack development quickly.

ILLA proudly announces a partnership with Hugging Face, a suite of natural language processing (NLP) tools and services. They are most well-known for their open-source NLP library, which provides text generation, language translation, and named entity recognition tools. With Hugging Face, ILLA is more productive than before. Our users can do more with AI.

ILLA Cloud provides dozens of commonly used front-end components, allowing you to build different front-end interfaces based on your specific needs quickly. At the same time, ILLA offers a connection to Hugging Face, allowing you to quickly connect to the API, send requests, and receive returned data. By connecting the API and front-end components, you can implement the requirement that users can enter content through the front end and submit it to the API. The API returns the generated content to be displayed on the front end.

You can check ILLA’s website here at: illacloud.com

GitHub page: github.com/illacloud/illa-builder

Join Discord community: discord.com/invite/illacloud