Is the influences futural AI bring to software engineers are that scary?

admin管理员组

文章数量:1597416

(Image credit: Baidu)

By Robin Li July 2021

---

“Artificial intelligence will undoubtedly affect the development of human society in the next 40 years.”said by Robin Li, founder of Baidu, on WAIC2021.

Article continues below

What does Robin Li mean? Thinking carefully about his words, that means the development of AI will bring a lot of changes to society in the future. As we can see, in the past several years, the development of artificial intelligence is quite rapid, but for this, some people are happy, and some people are sad. People rejoice in the progress of science and technology but also worry about the harm caused by the rapid development of science and technology, especially the AI technology.

Looking at the development of artificial intelligence over the years, it can be found that artificial intelligence affected by the industry is not in the minority, so many people worry that in the future artificial intelligence will gradually eliminate their own industry, and even some people think that artificial intelligence will eventually destroy human beings. It's normal to have such concerns, because no one wants the result of robots ruling the world in movies.

Some movies like this:

- 普罗米修斯 (豆瓣) (douban)

- 异形：契约 (豆瓣) (douban)

- 终结者 (豆瓣) (douban)

As a future software engineer, I was also worried about whether software engineers would lose their jobs, after all, there are now a lot of code writing assistance by artificial intelligence occur. But now I don't think much of it. “The growth of artificial intelligence is almost unstoppable, even the complex industries can be affected, but when considering these possible impacts, you will find that they are not that panic.” Simply speaking, in the future, the rapid development of AI is likely to change the work of the software engineering industry, but these changes are not worth our panic.

Artificial intelligence can write its own programs, are you panicked? (Credit: Baidu)

---

Cooperation will occur more and more in the future between AI and software engineering industry, which will be advantages for both developments. 

AI has many increasing abilities nowadays, according to artificial intelligence blogger, Kaikeba(2021), “The core increasing capabilities of artificial intelligence can be divided into three levels, called computing intelligence, perceived intelligence, and cognitive intelligence.”

 Computing intelligence, that is, the machine has super storage capacity and ultra-fast computing power, can be based on massive amounts of data for in-depth learning, using historical experience to guide the current environment. These increasing abilities may have a great cooperation of software engineering in the future. According to professor C.RichR. and C.Waters(2021), these may be reflected in several aspects, for example, the powerful ability of computing can let AI operating many times to make the code more rational, and in the future increasing perceived intelligence and cognitive intelligence may make AI working as a human which can perceive people's ideas correctly and write code as people want.

You may consider it as an incredibility, but nowadays, many software companies like Kite companies in Silicon Valley have used AI to assist their code writing, this AI can predict the code content you may write next step. And many users agree that such cooperation greatly reduce their coding time, leaving them with more time to think. On the other hand, it also improves the development of AI, because AI is composed by code in a sense.

But why cooperation will occur more and more in the future? In professor Yuanyi928(2020)’s view, it’s very hard for AI replace software engineering in the near future, to improve human’s life, the cooperation is a more realistic and executable. AI can make poems like human nowadays, because the code also like a language, many people think that Artificial intelligence can easily do the job of software engineers. “However, unlike natural languages, code is very sensitive to noise” according to professor Yuanyi928, that means natural languages can have errors, but code languages shouldn’t. Even if 99% of the software code is right, it can’t be compiled successful, the software requires there are no errors in code. So, the reason has two aspects, on the one hand, it is very hand for AI to replace software engineering in the several years, on the other hand, the cooperation has many advantages for both AI and software engineering.

Although artificial intelligence is still in the code-assisted phase, and there is still a long way to go before there is real cooperation, but whatever the smart code complement tool, TabNine, created by Jacob Jackson or the code translation conversion tool, Transcoder, created by Facebook, these simple assistances bring a lot of advantages. Generally speaking, in the future, these advantages will improve the cooperation between them, and this will become a good positive feedback which constantly improve both developments.

 The application of AI in code-assisted phase (Credit: Baidu)

---

If replacement occur in the future, a new job will bath for original software engineers-“artificial intelligence trainers”, which will not only keep their employment rate but also be important for development of AI.

Science and technology is the first productive force, and the constantly improving of the level of science and technology can enable mankind to live a better and more convenient life. There is an old saying: fall behind will get beaten. So nowadays, the competition of science and technology in various countries is becoming fiercer and fiercer, AI is no exception.

Throughout the ages and present, the human’s world is developing in a convenient direction. Thus, artificial intelligence seems to become an inevitable trend in the future. As you can see, some work is gradually being replaced by artificial intelligence today, is that means even software engineers will be replaced in the near future? A report by Michael Osborne and Carl Frey, a professor at Cambridge University, found that software engineers had an 8.5 per cent chance of being replaced by AI.

---

Although AI seems have no abilities to replace software engineers soon, but we can’t guarantee that AI won’t have huge throughout in next few years.” they said.

---

The “artificial intelligence trainer” (Credit: Baidu)

---

For my understanding, that means we can't guarantee that this won't happen. According to a report by the Oxford Institute for Economic Research(2019), “AI Robots will replace 20 million manufacturing jobs worldwide and eliminate a variety of low-skilled jobs over the next decade.” But the report also pointed out that the new industry will occur as much as the replaced, and these new occupations are generally closely related to the original traditional occupation. In their view, that is because, if one industry is replaced by AI, it should another industry to occur to support and adjust AI. That industry is called “artificial intelligence trainer” by Professor Wang Tingwen(2016) in Sina Technologies.

“Artificial intelligence trainer” can be an important work for the development of AI and original industry. Because, if AI has no supports from original industry, it will loss the develop direction. It needs many workers of original industry and also programmers. And if no new relative industry occurs, the original workers will be faced with unemployment. Thus, although artificial intelligence may be an inevitable trend in the future to replace traditional occupations, it’s not that harmful and bad on some extents. If software engineering industry is replaced by AI, it is no exception. Think about it, AI replaced you with a repetitive and tedious job which programming today and night but give you with a new job that only “training” the AI like an animal. That seems quite more interesting than the original work. It will provide software engineers a new direction related to their original job to develop. Generally speaking, although the chances of artificial intelligence replacing software engineers in the next few years are slim, that doesn't mean it won't surely happen. Although it happens, it is not something to panic about. It will occur a new job called “artificial intelligence trainer” for software engineers which is important for AI and original industry.

No matter how artificial intelligence develops, even if all industries are replaced, artificial intelligence trainers will not be replaced.

Some people may have different view about “artificial intelligence trainer”, they think the work of “artificial intelligence trainer” is temporary, in the end, it will still be replaced even disappeared in the future. However, in my view, no matter how artificial intelligence develops, even if all industries are replaced, artificial intelligence trainers will not be replaced. Because artificial intelligence trainers are too important for human’s future.

Hawking, the greatest scientist of the 21st century, left seven world-shaking predictions before his death. And one of the predictions is AI will replace human and domain the world. Although Facebook and Google are obsessed with artificial intelligence, Hawking didn’t trust it one bit. He claimed it could be the “worst invention of the history of our civilization.” Speaking to Wired UK, he said: “I fear that AI may replace humans altogether.” Bill Gates, Microsoft's founder and philanthropist, has also warned people about the dangers of artificial intelligence.

 AI robot destroyed humanity (Credit: Baidu)

---

Why do the world's most recognized "smartest people" think artificial intelligence could be a disaster? As for my view, that is because they are worried about that artificial intelligence is likely to outstrip humans in intelligence in the future. It seems incredible, can AI really be smart than human? However, their concerns are not unreasonable. Although the intelligence of AI seems develop slowly nowadays, once artificial intelligence's IQ outstrip a certain critical point, it will grow with a faster and faster speed which human can’t reach. More intuitively, AI's IQ development curve may be an exponential curve. It’s very very flat before a critical point, and when it reaches a critical point, it slowly begins to accelerate, faster and faster, until it approaches a vertical straight line. An exponential curve (Credit: Baidu)

---

The higher IQ of AI than human probably will cause AI above human, which means AI may beat human and domain the earth. As shown in the movie “The Matrix”, AI robot have a world war with humans. As I analyzed in main point 2, the development of AI is unavoidable, so, how could we stop AI from destroying humanity? As professor Wang mentioned that we need “artificial trainer”, who control AI to avoid potential threats from AI to human. We can compare artificial intelligence as a circus lion, think about it, if no one trains the lion, it will hurt humans even kill humans, and so can artificial intelligence, we need trainers to train them, not only “teach” them knowledges making them useful for human, but also limit their behaviors letting it no threat to human. Generally speaking, AI trainer’s work is very important which must not be replaced, because it’s about human’s survival.

o conclude, in the future, the increasing ability of AI will bring changes to software engineering industry, but these changes are not worth our panic. Because one of the changes may be that cooperation will occur more and more between AI and software engineering industry, which will be advantages for both developments. If we think worse, the change is AI really replace software engineering. But it’s not that harmful and bad on some extents, a new job will bath for original software engineers-“artificial intelligence trainers”, which will not only keep their employment rate but also be important for development of AI. And it’s not necessary for us to worried that even artificial intelligence trainer will be replaced, because the trainer’s work is about human’s survival which must not be replaced.

Perhaps Hawking's prediction really came true in the future, and humanity is facing extinction because of artificial intelligence robots, which may be due to the improper management of artificial intelligence by humans, or it may be that artificial intelligence has developed to a level that humans cannot manage. But can we abandon the development of artificial intelligence because of such concerns? People were afraid that the development of the automotive industry would bring pollution to the environment, so they opposed it, but today's cars seem to be indispensable in most people's lives. Things are always two-sided, and we can't always think about the bad side of it, right? 

本文标签： AIfuturalinfluencesBringscary