Artificial intelligence (AI) has undergone more than 60 years of development since its initial conception. In recent years, driven by new technologies such as big data and the Internet of Things, AI has achieved leapfrog development and entered its third golden age.
However, the rise of any new technology seems to trigger a series of debates, and artificial intelligence is no exception.
Since the emergence of artificial intelligence, various "futuristic theories" have begun to appear.
Some say that the current state of artificial intelligence is just a bubble that will soon burst. Others believe that artificial intelligence will threaten our jobs and even society.
Stephen Hawking once said, "In my lifetime, I have witnessed profound changes in society. The most profound of these, and the one with the most increasing impact on humanity, is the rise of artificial intelligence. The rise of artificial intelligence will either be the best thing in human history or the end of human civilization."
However, some people disagree, arguing that the view that "artificial intelligence threatens society" is far-fetched and that such concerns about artificial intelligence are completely unnecessary, like worrying about "overpopulation on Mars."
Today's artificial intelligence primarily assists humans with simple tasks, such as filtering spam emails and predicting the weather to provide drivers with directions. What's truly causing concern is a different kind of artificial intelligence called AGI, or Artificial General Intelligence. This is a form of AI that is as intelligent and flexible as humans, capable of solving problems that are insurmountable for humans.
Why do we have such a big disagreement about artificial intelligence? Why is there such a large gap between the views of the two opposing sides?
The following five questions may provide answers.
How far are we from unsupervised learners?
Currently, what we can achieve is only a narrow definition of artificial intelligence, which requires meticulously prepared datasets and manual training by humans. For example, to teach AI to recognize cat pictures, you need a massive database of photos, each manually labeled as either "cat" or "non-cat." Even if AI can accurately identify cats like a human, it will only be able to recognize cats; to learn how to recognize dogs, it would have to start from scratch.
How fast is AGI progressing?
AGI is the main reason for people's panic about artificial intelligence. Those who panic believe that we will build an unsupervised learner that will recursively evolve itself at a speed that humans cannot control. Day after day, year after year, AGI will become smarter and smarter, becoming an unsurpassable intelligent being.
The other side of the debate argued that this was merely a fantasy, akin to science fiction, and that such predictions about AGI were a misunderstanding of human intelligence.
How complex is intelligence?
This question leads us to the third point of contention. Those who believe the threat of AGI is imminent oversimplify the issue of intelligence. Newton believed that understanding motion only required three simple laws; similarly, electricity and magnetism can be understood and controlled through a few simple laws. Therefore, we might wonder, is intelligence similar? Can we also grasp and control intelligence through some fundamental principles? Those who worry about the threat of AGI might believe the answer is "yes," and that finding these principles would allow AGI to escape the confines of humanity.
However, the opposing view argues that the hundreds of unique functions of the human brain suggest that our intelligence comprises countless extremely clever evolutionary hackers, each of whom performs a great deal of understanding and organization work before we transmit knowledge to machines.
How special are humans?
AGI is a form of artificial intelligence that is smarter and more versatile than humans. Those who believe AGI is imminent often underestimate its potential. They fail to see the mysteries of human creativity. Conversely, another group sees the full extent of human potential, from intuition and emotion to consciousness itself. Humans cannot even comprehend how they perform their daily tasks, let alone create robots to do them.
Have we already started building AGI?
As mentioned earlier, the artificial intelligence we have today is called narrow AI because it can only do one specific thing. Is narrow AI the first step in building AGI? Many people confused about AGI believe that researchers are currently using the same technology to piece together these fragmented narrow AIs.
The other camp argues that artificial intelligence in the narrow sense and AGI are two completely different concepts and cannot be merged. Although they both contain the words "artificial" and "intelligent" in their names, their similarities are limited to this.
These are the five problems that artificial intelligence will face in the future.
Interestingly, the debate about automation and employment mirrors this sentiment. Those on the AGI side argue that with artificial intelligence, robots and automation will soon take over all jobs, threatening human unemployment. The other side, however, believes that automation and AI will not directly compete with humans. Instead, they argue that AI will not only increase human productivity but also create better, higher-paying jobs.
Undoubtedly, we are navigating uncharted waters. But one thing is certain: all these problems are harbingers of the future. As we struggle to cope with digitalization and intelligentization, our lives become increasingly precarious.
Disclaimer: This article is a reprint. If there are any copyright issues, please contact us promptly for deletion (QQ: 2737591964). We apologize for any inconvenience.
