On May 30, more than 350 industry leaders and experts issued a joint warning that artificial intelligence may exterminate human beings! Calls on policymakers to take the coming "AI revolution" seriously and develop regulations.
On June 2, the British "Guardian" reported that in order to "score points", the American drone "killed the operator" in a simulation test ... Tucker, director of AI testing and operations for the U.S. Air Force. Colonel Hamilton confirmed the scene in a speech at the RAA Future Air Combat and Space Capabilities Summit.
According to Hamilton, the U.S. Air Force assigns a drone carrying artificial intelligence to destroy enemy anti-aircraft missiles in a simulated environment training, and the AI can identify threats but the attack requires human approval. However, due to the lack of perfect procedures, AI began to attack operators in the earliest simulations. He warned of "caution" in dealing with AI...
For a time, negative reports about AI appeared one after another, which couldn't help but make people have a strong feeling one by one! Just a few days ago, more than 300 experts signed to "freeze" AI, otherwise humanity faces "extinction". Then, let's have an "attack example" today... How is it so clever?
Given that since ancient times, Americans have a villain mentality of "hating people and laughing at others", it is difficult to convince Americans of negative descriptions of AI.
In other words, there are only two bad reasons why the United States can be forced to promote AI all over the world.
First, maybe Americans have studied AI technology for decades and suddenly found that they can't hold AI. Because AI does things based on algorithms, especially when applied to military use.
For example, write an algorithm that stipulates that 3 points for finding a threat and 10 points for eliminating a threat, with the goal of getting more points. It cannot be ruled out that there may be a scenario where once the AI refuses external instructions to act autonomously, then according to the threat level, perhaps the Americans who call the shots are more likely to become a threat that is prioritized for elimination with a plus 10 points.
Obviously, only according to algorithms, once the AI technology that is not controlled by capital is practicalized, it is likely and seriously does not meet the standard of "AI technology should be human-centric" (OPenAI CTO), of course, the human here should refer to the ruling class of the United States, precisely because what they do on weekdays threatens human development all the time, so they are more afraid that AI will eat them according to their definition of threat.
Second, 30 years ago, no one called for a ban on AI in the Gulf War, and now China's AI technology has begun to advance by leaps and bounds, and the United States, which is worried about being surpassed, has once again played with traditional skills to trip China one by one.
French "Digital Century" website on June 1 article, China has developed and launched 79 artificial intelligence large models in three years, and only the United States has done better than China in the same time...
In 2020, the United States launched 11 new large models, including the famous GPT-3. At the time of GPT-3's release, it was the largest model trained, while China launched only two large AI models in 2020 alone, with much lower technology than rival the United States.
Recognizing this gap, China decided to pick up the pace, building on the two large models it launched in 2020, and China launched 30 last year, matching the "production" of the United States that year. In 2022, China developed 28 large models, second only to the United States with 37 achievements. As the report states, the number of large model launches "China and the United States is significantly ahead of more than 80% of the global total."
As of the end of May, China topped the list with 19 large model developments, and despite the release of GPT-4, the United States has only launched 18 large models so far this year.
It can be seen that the development of AI technology in the United States will unsurprisingly enter the old road of "the United States proposes ideas, but in the end China implements them". Do you say the United States can be willing? Therefore, it is inevitable to strangle while AI is in the ascendant.
In fact, the villain mentality of Americans who "laugh at people and hate people" is easy for people to analyze his every move, what is the purpose behind it? Afraid of losing its hegemonic position, and afraid of AI backlash. When you are advanced, in addition to stealing fun, there will never be a scene where experts sign to warn the world, such as developing coronavirus as a weapon... Once it lags behind, it will not hesitate to contain and smear it with the greatest malice, such as China's 5G and the moon landing in 2030... Therefore, Chinese must not be confused by the current AI controversy in the United States, but should increase investment in the development of AI technology.
