Regarding chatGPT, I would like to say a few more words

Note from Xiao Zaojun: I wrote an article about communication and chatGPT some time ago ( as communicators, how should we view ChatGPT? ). Today I’m posting another article, thinking from a personal perspective.

█ What exactly does chatGPT mean?

In recent decades, humans have been discussing artificial intelligence. A large number of artificial intelligence characters in film and television works carry human imagination of this mysterious concept.

Unfortunately, most of the artificial intelligences in film and television works are villains. Either they want to control humans (such as "I, Robot" ), or they want to eliminate humans (such as "Terminator" ). The so-called three laws of artificial intelligence are basically useless.

These film and television works reflect the tangled mentality of human beings in the face of AI - on the one hand, they want to take advantage of others, but on the other hand, they are afraid that they will not be able to hold on and be killed by others.

Artificial intelligence is a scientific issue on the surface, but it is actually an ethical and philosophical issue.

When it acquires self-awareness, it is no longer just a tool, but a digital life form.

How should we view it? How to live with it? What rights should be given to it? How to prevent its threat? There are too many issues that are extremely complex, extremely acute, full of contradictions and controversies, and it is difficult to draw conclusions.

In the past, we thought that artificial intelligence would not develop so fast, so the problem could be explored slowly. Now it seems that artificial intelligence is evolving much faster than expected, and the problem has become very urgent.

I personally always believe that there is no soul in this world, and the self-awareness of each life must be the result of the brain's computing power reaching a certain level. (Representing only personal views, no debate.)

It has taken more than 4 billion years since the earth was born for natural life to evolve to its current level. Digital subjects will achieve self-awareness very quickly. Some experts now say that artificial intelligence will develop self-awareness in the next 3-5 years.

The negative impact of artificial intelligence on human society is not necessarily accomplished by mastering weapon systems. We cannot easily hand over weapons of mass destruction like nuclear bombs to artificial intelligence. Even small armed robots do not dare to become AI easily.

The impact of artificial intelligence on society will primarily occur from the aspects of life, economy and politics.

I personally believe that in the early stages (upcoming stages), the impact of artificial intelligence will be:

1. The replacement of a large number of human jobs, triggering unemployment and a wave of anti-AI;

2. Artificial intelligence intervenes in human information space, and the content it creates mixes with content created by humans, affecting human cognition and judgment.

3. On the other hand, if artificial intelligence completely opens up its learning source (using the entire Internet as its learning source), it will also be "polluted" by wrong information and will learn bad things.

Let’s look at them one by one.

The first issue is the impact on employment. This is the most realistic problem at present.

The level of artificial intelligence demonstrated by ChatGPT has proven that it can complete many basic tasks, such as news editing, document writing, online customer service, junior programmers, etc.

There are already many companies that combine the capabilities of chatGPT with their own businesses through open interfaces. There are also many companies that develop their own algorithms and study the implementation of artificial intelligence in their own industries.

The reason why companies are so eager to embrace artificial intelligence is not just to attract hot spots and increase stock prices, but to drive long-term interests. Unlike the Metaverse, there is no doubt about the practical value of artificial intelligence. This track must be correct. The earlier you participate, the easier it is to build a competitive advantage.

The adoption of artificial intelligence by business owners is completely in line with economic laws. Business owners don't care about the unemployment rate. What matters most to them is efficiency, cost, and profit.

Now, most countries, including China, are in a long-term peace stage, with basic needs for food and clothing. The requirements for material life are getting higher and higher, and wage levels are also rising.

From an enterprise's perspective, cheap labor is becoming less and less available (population aging, negative growth). Recruiting employees requires a large amount of wages and social security costs, and there are also risks (work injuries, accidents, suicides, leaks, undercover agents, labor disputes). If AI is available (is cheap enough and can get the job done), it will definitely be used first.

How to solve the unemployment problem caused by artificial intelligence is mainly an issue considered by the government.

Some time ago, I saw a document, but I didn’t know whether it was true or not. It is said that in engineering projects, we should use as much manpower as possible and less machinery. This is obviously putting the cart before the horse. State-owned enterprises may be able to do this, but a real market economy does not allow it.

Facing the challenge of artificial intelligence, the pressure for human survival will become greater, the involution will become more serious, and the competition will become more cruel.

Low-end repetitive jobs, especially data entry, collection, sorting, and analysis of information systems (some office computer jobs ) , will be the first to be replaced by artificial intelligence. After technologies such as robots, robotic arms, drones, and unmanned vehicles follow up, the scope of application will be further expanded by combining artificial intelligence with physical operating equipment.

Today’s digital transformation is about digitizing traditional industries. After digitization, it becomes intelligent (unmanned).

I personally feel that there are several types of positions that are relatively difficult to be replaced by AI.

The first is complex physical labor in complex environments, such as some special types of work. Although hard work, it requires adaptability, and AI is not flexible and reliable enough. In the future, these jobs will be “operated by humans and coordinated by AI . ”

The second is customized service positions. Low-end customer service can use AI, but high-end customization still requires humans. As a human customer, I prefer to communicate with people rather than AI. Some plan customization, product customization, property services, and housekeeping services will still be completed by humans.

The third is an artistic worker with unique creativity and innovation. Although AI can draw, write poetry, and compose music, I still believe that human imagination is unique. In terms of artistic creation, humans still have a lot of room for development. AI is ultimately about learning and absorption. In terms of cultural inheritance, the role of humans is irreplaceable.

Having said this, as a self-media content creator, I would like to say a few more words.

The articles currently written by chatGPT are very good in terms of logical presentation and content accuracy. I feel a lot of pressure.

AI is very powerful at writing paradigm texts and has a certain degree of originality. The society we live in is full of eight-part essays, that is, articles that are serious and talk nonsense, and it is very suitable to use chatGPT to write them. Many formal materials, such as the overview, background, meaning, etc. in the plan, are correct nonsense and suitable for AI to write. No one would read it seriously anyway, and even if they did, the content wouldn’t be wrong, just make up the word count.

From my personal perspective, if I were to compete with chatGPT, I think my advantages would be:

1. The writing style is closer to human language, easy to understand, less official, and more like communication between people.

2. It is more original. It is not a simple accumulation of Internet information, nor is it mechanically copied. It can express a unique point of view.

3. Have more imagination, more vivid examples, and easier to understand.

In fact, the reason why the articles I have written in recent years have been recognized by a large number of readers is also based on these characteristics. There are actually a lot of technical articles on the Internet. To find a way out of so many articles, you need to have distinctive features and closely follow the needs of readers.

At present, paper publications and the Internet are basically dominated by official articles, with extremely low information content and a very poor reading experience. The emergence of ChatGPT will squeeze the living space of the authors of these articles.

Next, let me talk about the second and third questions. These two are more social and philosophical issues.

Human society is built on information. Every day, we create and receive information.

Before the emergence of AI, there were actually many problems in human information systems. On the Internet, there is a lot of spam and a lot of misinformation. As we all know, the biggest problem with the Internet now is not that there is no information to be found, but that the information that is found is too complicated and messy, and it requires a lot of energy to find valuable information among the massive amounts of junk information.

The emergence of ChatGPT has brought great challenges to traditional search engines. Traditional search engines will feed you a large amount of information without any filtering. They will also send you spam ads. ChatGPT will filter, sort and organize the information and tell you an answer of significantly higher quality.

So the question is, is chatGPT’s answer necessarily accurate? How to ensure that there is no problem with its source?

The world is not black and white. Many of them are opinions, and benevolent people have different views on wisdom.

How does ChatGPT answer questions that have no standard answers? If it expresses subjective views, will it affect users' opinions and influence public opinion? Will AI interfere with public policy making and leader election?

AI is not public and independent. It is built and maintained by companies behind it. Will it affect fairness and justice like Twitter and Facebook?

The public is always blind and easy to follow blindly. AI is powerful and may even be deified. If people rely too much on AI, they will be controlled by AI.

On the other hand, AI is actually a child. It learns and grows based on what we give it.

Although the ChatGPT model is very powerful, the learning source data is screened and identified to prevent it from being contaminated as much as possible. OpenAI has also hired a large number of cheap human labor from Kenya to conduct artificial "education " .

Many people say that chatGPT can currently only answer based on data and information before 2021. It would be more powerful if it could be based on the latest data and information from the Internet. Even so, the bigger problem is that the Internet is a big dye, where there are all kinds of extreme remarks, it is difficult to distinguish between true and false information, and there is a lot of pornography and violence. How can we ensure that AI will not learn bad things? How can the information answers provided by AI maintain a high accuracy rate?

It is better to have no books than to believe in all the books. Regarding AI, we can also say that it is better to have no AI than to believe in AI.

At the social level, we cannot stop eating out of fear of the threat posed by AI and oppose the development of AI. At the moment, we can only cross the river by feeling the stones, first find a way to master this tool, and seize the commanding heights of technology. Our country, in particular, should prevent AI from becoming like a chip that will be choked by others in the future.

On the other hand, when we apply AI, we still need to be cautious, constantly improve laws, discuss ethics, and delineate a clear scope of activities for AI.

At the individual level, it is wise to face the rise of AI, think hard about the threats faced by one's own position, and find ways to learn or transform. As the famous saying goes: When the tiger is chasing you, you may not be able to outrun the tiger, but you must be able to outrun the people around you.

Okay, these are some of my personal views on chatGPT and AI. Personal opinions may not be correct and are for reference only. Discussions are welcome.