Author: Duan Weiwen
Source: Xinjing News
Since its launch on November 30, 2022, the various performances of the new generation of artificial intelligence chat robot ChatGPT are very stunning.
From continuous answering questions, generating abstracts, translation documents, to information classification, writing code, screenwriter, doing homework and writing papers, ChatGPT can almost cope with it.Moreover, ChatGPT will question your prerequisites and even refuse to ask for improper request.
Overnight, people who have registered their accounts have encountered machine cats, pulling ChatGPT and asking.Many people are excited for this, but many people have experienced the shock of the future from it. Worried about whether they would wake up, they would be omnipotented by the omnipotent universal artificial intelligence like the nobility of the Middle Ages.Crossing.
You need to see the disruptive effect of ChatGPT
ChatGPT was born, unlike the shock waves of ALPHAGO defeated human Go players five or six years ago. This time, people's heated discussion is no longer just a lot of worry about it.The panic of "change the rules of the game".
It is not difficult to imagine that in the near future, the company manager can convert a rough document into an elegant presentation without just pressing the button or sending a instruction to the voice robot without having to worry about the company's secrets.This undoubtedly subverts our work scenario that we are already familiar with.
Professor of MIT and Cosmologist Tig Mark, a topographic diagram of human capacity in his life 3.0, believes that writing and design belong to the final highland of human intelligence.The changes that ChatGPT may bring may mean that the water level of artificial intelligence may be drowned first, that is, the human high slopes such as writing and design.
In fact, whether this time people worry about ChatGPT or the loss of the scientific god Tig Mark, there is a reason for the regularity of breakthrough technology development.
Simply put, because of the scale effect of the large models of pre -training, ChatGPT and similar software can predict the next word in all sentences, thereby emerging the incredible ability of connecting words.
One important reason why people are shocked by the breakthrough progress of ChatGPT is that people do not believe this kind of technical route that patch out language at first.But there is no doubt that ChatGPT does have a good performance in the fluent language of grammar.
As a result, it is not difficult to get a lesson that many of the routes of subversive technologies do not necessarily meet people's consistent and advance cognition.
In the past 10 years, from deep learning to the development of genetic artificial intelligence, especially this point -what intelligence should be for artificial intelligence, people’s general cognition and artificial intelligence breakthrough innovation are delivered.There is a gap between forms.
Therefore, when it comes to the attitude of generating artificial intelligence such as ChatGPT, regardless of whether or not they are aware in advance, no matter where they are in and the track, they should see the disruptive effects brought by it.In other words, we must consider it as the engineering engine that leads the future, and actively embrace this wave of artificial intelligence gold rush.
Humans have the ability to clarify ChatGPT governance ideas
Like all major technical breakthroughs in human beings, the emergence of generating artificial intelligence such as ChatGPT and its future, in fact, rooted in the overall technical and social system, will inevitably be based on our imagination of its possible changes.
In these new system construction and future imagination, some are relatively neutral changes, while others have positive changes to everyone.
For example, it may not take long for people to watch videos of any language with the help of general artificial intelligence translation tools, and people who use different languages can also communicate smoothly with each other.
But as people have seen recently, ChatGPT's virus epidemic has also caused a lot of criticism.From the concerns of the teacher's academic integrity of the student's thesis, to the artist's anger imitated by the artist's work without knowing it, let alone make fake news and promote public opinion orientation. The abuse of social robots will inevitably become more and more intense.
Therefore, some researchers are worried that in the new wave of innovation competitions of generating artificial intelligence such as ChatGPT, will they step on the rear of search engines, social media and big data business, so that ethics such as integrity, privacy, intellectual property rights and other ethics and other ethics and other ethics and intellectual property rightsLegal rights and basic values have become victims.
How to effectively take this problem?In the past, the thinking of digital governance and artificial intelligence governance still applied under the wave of ChatGPT?
It must be pointed out that the essence of generating artificial intelligence such as ChatGPT is the revolution of the knowledge production method. It is not only a redefinition of knowledge, learning and creation, but also a re -understanding of human intelligence and machine intelligence.
In the study of massive text data sets, the prediction and related knowledge formed by ChatGPT is the essence of automated medium that connects the knowledge of human beings together.
From this, in the tide of ChatGPT, when we are preparing to go up, it is necessary to think deeply about related issues first, and we have the ability to clarify ideas and come up with the response.
Development should emphasize border thinking and change of wisdom
ChatGPT, a knowledge production method based on human group knowledge and memory, will inevitably conflict with existing knowledge production methods based on individual knowledge and memory.ChatGPT's high dependence on massive data sets and other data will inevitably bring conflicts in the concept of values that are difficult to resolve during its patchwork knowledge structure.
Therefore, we should try to prevent extremely bad situations, but use the general information review mechanism carefully.In view of the artistic works and code sets involved in ChatGPT, it will inevitably conflict with the original intellectual property rights and other rights, and it will also need to establish a new social contract and legal system.
ChatGPT's current success is mainly technical. While we embrace the opportunities brought by it, we must further ask this technology's impact on the connotation of human knowledge and intelligence and machine knowledge and intelligence.
Similar to the relevant knowledge of deep learning and intelligence, this productive knowledge and intelligence are predictive, and its predictability is based on existing knowledge.
It can also be seen from this that while we embrace such machine intelligence, we should see the current unique reasoning, reflection and creativity of human intelligence, which should also be the direction of human efforts to enhance todayEssence
In the face of the huge dividends of genetic artificial intelligence such as ChatGPT and coexist with unprecedented disadvantages, it is also necessary to formulate corresponding policies at the speed of technological innovation and simultaneously introduce dynamic governance mechanisms in ethics, law, and education.
In this process, the most important thing is the innovation of real -time policy and mechanism, especially the emphasis on border thinking and changing wisdom.
On the one hand, grasping the worst issues, such as fraud and other crimes to curb, at the same time, we must realize that massive text data sets are becoming a general knowledge production element. In terms of content reviewThe potential of knowledge production.
On the other hand, we must also consider the transitional strategy of new and old knowledge production methods. Related policies and governance mechanisms should be committed to reconciled knowledge productionThe friction brought by method conversion.
For example, in the field of education, a policy that is relatively isolated from the basic education stage from automatic writing technology can be adopted to promote the development of young people's expression and creativity.
The generated artificial intelligence seems to have passed Turing test, but the uniqueness of human intelligence is that it can change with the changes of the environment and machines.Therefore, the Turing test is not a lifelong college entrance examination.This is also because we can always move the Turing boundary between people and machine intelligence.
In short, in the face of the sudden new species of ChatGPT, we should not easily fall into its "beautiful new world", but use the unique wisdom of humans to let it fly with us!
