The Code in Your Hand May Become Their Bullet

Kim Jong-un, 34, and Trump, 72, representing North Korea and the United States, which have been hostile to each other for nearly 70 years, held hands for about 13 seconds at the Capella Hotel in Singapore.

North Korea's nuclear program in recent years, the U.S. military's proactive offensive, Russia and China's defensive attitude, and the volatile international situation have all caused the entire world to fall into a period of anxiety.

Unlike the Cold War decades ago, contemporary countries are increasingly focusing on investment in information technology to create more advanced military weapons and cultivate more and better military talents to strive for their own world status.

When it comes to developing military technology, there is no need to say much about the United States, as they have always advocated talent first.

However, the trend in the United States is that mainstream well-known institutions such as MIT and Stanford have trained a lot of engineering talents, but they are usually applied talents who go directly into the industrial world, and rarely are they trained for national defense and military purposes.

Unlike other countries, my country has a comprehensive focus on military science and technology. For example, well-known universities such as the University of Science and Technology of China, the National University of Defense Technology, and the PLA Information Engineering University all have modern science and technology majors such as computer science, electronic information, and big data, and almost all provide targeted training and delivery for national defense and military affairs.

In addition, our country has a type of students with Chinese characteristics: national defense students, who appear in various universities and enjoy preferential tuition policies and strict training programs.

Our old neighbor Russia also attaches great importance to talent cultivation, especially in the field of computer science. Everyone knows that their mathematical foundation is so good, so since ACM became an international competition in 1977, Russia has basically been the country with the most championships.

The fighting nation is incomparable

I believe many of you have watched this year's ACM held at Peking University. Several Eastern European universities, including Russia, once again made their presence felt and severely outperformed other countries.

However, due to the lack of an entrepreneurial environment and too many monopolistic industries, it is difficult for these smart programmers to develop in the field of science and technology. If they want to contribute to the country, they can only build tanks.

North Korea, led by the great leader General Kim Jong-un, who was all over the news today, is in a rather awkward situation in the talent competition. There are only a handful of schools in the country that can become universities. In the field of computer science, only Kim Chaek University of Technology can come up with something good, so it can be said that there is a shortage of talent.

In addition, the entire country is relatively closed. After the nuclear program was aborted, these few science and engineering talents can only continue to build submarines for the country. I am also very worried about North Korea's technical talents. After all, the great leader General Kim Jong-un has boasted too much.

After more than three months and more than 3,000 employees protested together, Google finally gave up its plan to renew its contract with the U.S. military for Project Maven. This is a technological victory, but it is by no means the end of the cooperation between Internet companies and national military programs.

The Project Maven project was first revealed by an internal employee. At that time, Google was providing AI knowledge and technical support to the US military through a private intermediary company, helping the US military develop drone AI systems, allowing the military to use Google Cloud and guiding and supporting them in using Tensorflow.

Although it has caused unprecedented controversy within the company and aroused great public pressure, it has only stopped the project cooperation with Maven. If you think that Google has stopped its involvement in the US military operation, it is too naive.

The U.S. Department of Defense invests nearly $10 billion in information technology research each year, which is much higher than the total defense investment of other countries, or even dozens of countries combined. And Project Maven, as we know it, is just the tip of the iceberg.

In addition to Google, IBM, Microsoft, etc. are also providing technical support for various military projects of the US Department of Defense.

Engineers from Internet companies have been promoting the development of AI technology. Can such civilian airliners be directly used to develop military weapons and participate in military wars?

Judging from the technologies that have been realized so far, the autonomy of AI is based on historical data. Relying on powerful computing power, AI can make simple predictions about future behavior based on history.

But these data are not enough to support AI to form a moral code like humans. They don’t know racism or the meaning of killing. They only mechanically execute the commands of the established algorithms. Therefore, over-reliance on data is also a fatal flaw of AI.

If the input data is negative or even biased, AI will behave in a biased manner and even amplify bias, such as racial discrimination, sexism, etc. Humans can restrain themselves with morality, but for AI, data is its morality.

According to AI's logic, Trump's airstrike on Syria would not be as simple as launching a few missiles. Moreover, under human control, it is difficult to ensure that the input data is not subjective, and this subjectivity is biased.

Google's withdrawal from a project with the military is a victory for public opinion, but it does not mean that the AI technology developed by Internet companies will not continue to be applied to the military. The British government recently issued a set of AI development principles:

• AI should be developed for the common good of humanity;

• AI must operate under the principles of understandability and fairness;

• AI must not be used to undermine the data and privacy rights of individuals, families or communities;

• Every citizen has the right to an education that enables them to thrive mentally, emotionally, and economically alongside AI;

• AI should not have the autonomous power to harm, destroy or deceive humans

Dear AI engineers, please read this article based on your own projects.