Editor's note: Time does not go fast to the end of the year, and 2023 is about to come to an end. We always sail on the river of time, conquer the waves, and linger for the scenery along the coast. The years flow silently, leaving marks of different shades in our hearts. LIZHI Rui Review Year-end Special Plan - Look back at 2023 with you, pick up the shining memories of the year, and go on a new journey.
Wen Zhou Wei
From forgetting to write to no AI is unspeakable
This year, AI tools such as ChatGPT and Wenxin Yiyan have exploded, which is a microcosm of AI artificial intelligence, and the most touched and touched changes this year, changes in life and work, impacts, etc., can you provide some perspectives?"When I received this request, my first reaction was to enter the text in the dialog box of an AI software.
It seems to have become a habit and even a dependence on me. 2023, from the first contact to deep dependence, is my first year of AI. I also believe that not only me, but also many friends who have a lot of text work needs, once they use AI auxiliary tools, look at the large paragraphs of text that pop up in the dialog box, and what kind of pleasure and relief it used to take time and effort to type. The thrill of this technological leap is also very addictive. Even, many people have reached the point where it is difficult to speak without AI.
The last time we entered the information age, when keyboards and screens replaced pen and paper, the skill of writing deteriorated unconsciously, and many Chinese characters could no longer be written stroke by stroke. It can be said that the price of pursuing efficient text output is that the traditional writing mode of our hand-eye-palm cooperation has been greatly changed. This time, when we are trying to cede more of the right to "think" in exchange for convenience and efficiency, what awaits us?
The decision-making process of AI is a black box that no one understands
Artificial intelligence is not by design, but by training. The gpt-3 I used5 The knowledge base is based on all network data as of January 2022. Every decision it makes is related to this data, and that's the problem. No one, including scientists, knows how AI makes decisions, it's a complete black box.
When a traditional program goes wrong, engineers know what the problem is, and they know how to fix it. The problem with AI is in the data, not in the program, and it needs to read information from all over the web to do so. We can modify the program, but we don't know how to modify the data, because that's a fact that already exists on the web. No one knows what piece of data the AI is affected by, and how to modify itWhat's even more frightening is that all the negative data on the web is also absorbed by the AI at the same time, including discrimination, hate, bias, and lies – the impact of which is difficult to measure and repair.
AI seems to understand everything, but it doesn't really understand because it doesn't need to understand. The reason why AI can give the answer is that when A occurs, B has a certain probability of occurring when it occurs in the data. It doesn't care about the reason behind it, and it doesn't understand the meaning of each character. For it, it's all about numbers and the relationships between numbers. It's similar to animals, right?Animal instincts are evolved, and evolution only focuses on the result without investigating the cause, and the right survives and the wrong extinction.
The reason why human beings are different from other animals and AI is precisely because we have education, we learn and evolve by thinking. The same question, through in-depth research to get the answer, and ask the AI to get the answer, even if the result is the same, the process is bound to be different. The value of human learning lies in the process, not the answer. However, the advent of AI has disrupted this process, and humans are beginning to be able to get answers without having to understand why, which means that in the future, humans can stop learning and rely entirely on AI.
At the end of May 2023, 350 heavyweight AI scientists and industry players jointly issued a public statement on AI risk, calling on the world to face up to the threat of AI to human survival. The concern that humans will become inferior citizens is that they are so dependent on machines that no one wants to be educated, and that understanding, thinking, and exploration will be of no value to humans, and machines will become the real masters of the world. This is shown in many literary works. For example, in the animated movie "Robot Story", all the work of the Ark spaceship that humans take to escape from the earth is done by robots, and humans are only responsible for "living happily". But, hundreds of years later, things have changed ......
Pay attention to the "4C" and not be eliminated by the era of artificial intelligence
The pursuit of physical and mental emancipation brought about by technological progress is a "double-edged sword" in the course of human evolution. "New"It doesn't necessarily represent"Okay", but thinks"New"Necessarily represented"Not good"It is also a misconception. Pessimism about science and technology is even more detrimental to the development of our civilization. In his other book, A Brief History of Today, Yuval Harari, author of A Brief History of Time, argues that the focus of education in the future should be on the "4Cs": Critical Thinking, Communication, Collaboration, and Creativity. To put it more broadly, education should not focus too much on specific job skills, but on generic life skills. The most important thing is to be able to adapt to new things, learn new things, and maintain mental balance in unfamiliar environments. To keep up with the world in 2050, humanity will not only have to invent new ideas and products, but most importantly will have to reinvent itself again and again.
People should still dream boldly and walk in the forefront of science and technology
We can't say for sure what the new technology will change, but I have full confidence in the young people who grow up with it. I believe they will find new ways to use technology, just as our generation used computers and satellites, and each generation uses technology to realize human dreams.
As long as the more people believe that the technology of the future is positive, the more likely it is that the future will be positive. We will choose what the future will look like through our beliefs and actions. The race with AI has begun, and no one is a bystander. The game has started and we are all on the board. If you want to win, you can only dream bigger and think deeper.
It's not a choice between utopia and dystopia, nor is it something else we're up against. We must give full play to our ambitions in order to stay ahead of technology. We're good at teaching machines to do our job, and we're only going to get better at it in the future. The only solution is to create new jobs, new tasks, and new industries that we don't even know how to do.
Technology can make us more human, because technology gives us more freedom to be creative, but humanity is not only about creativity. We have other qualities that machines can't match: they only have instructions, but we have goals. Machines don't dream, not even in sleep mode;Humans dream. We need our smart machines to turn our biggest dreams into reality. If we stop dreaming, if we stop seeking greater goals, then we are just machines.