Remarks by He Junke, Executive President of CAST, at the Third Scientific Culture Forum

Peking University, Beijing, December 18, 2023

Currently, artificial intelligence (AI) technology is flourishing and industrial integration has broad prospects. The technological revolution is profoundly changing the production modes and lifestyles of human beings. The scientific culture in the AI era is evidently a global topic that spans politics, economics, culture, and the society. Faced with such a profound and widespread change, various forms of inadaptation are inevitable. How will AI change industrial economy and research paradigms? And how will it influence social governance and ethical norms? The unknowns outweigh the knowns now. In the face of the new stuff, some people eagerly embrace it; some are scared; and some have prejudices. Just like the situation before previous technological revolutions, people have intense discussions and divergent opinions. However, will a simple repetition of past scenarios happen this time? No one knows! It’s in pristine condition and we anticipate experts and scholars from all fields can offer a preliminary discussion framework.

How will the scientific culture in the AI era differ from that in the agricultural civilization era, industrialization era, and the information era? What are we perplexing on earth?

Firstly, how will AI affect the substantiality of scientific culture? Substantiality is the primary feature of scientific culture. Earlier forms of so-called AI in the primary stage were essentially reactive machines and memory reproduction, which seemed belonging to automation. They can be considered as applications based on objective reality. The generative AI used neural network algorithms and large model technologies to enable machines to "create" extensive new logic based on human wisdom. In the future, advanced and super AI are more prone to "create" new logic based on human wisdom. We cannot define whether it is real or illusionary. The connotation of scientific culture is faced with new challenges.

Secondly, how will AI affect the rigor of scientific culture? The rigor of the scientific and technological research paradigm is the defining characteristic of scientific culture. The traditional scientific and technological research is deemed convincing primarily because it relies on rigorous causal logic. The most distinctive feature of AI technology is the widespread application of new scientific research paradigms. For example, we can conveniently know what to do in most cases through correlation analysis based on big data, but we cannot define what the "correct" approach should be. How should we define correctness?

Thirdly, how will artificial intelligence affect the inclusiveness of scientific culture? We have acknowledged that science is a universal tool driving human progress, and also the importance of ensuring fair and open access to scientific knowledge and skills for everyone, regardless of wealth or status. Therefore, the scientific culture advocates inclusiveness, universality and openness, considering the overall values of human civilization. AI is undoubtedly a powerful tool for productivity progress. With its deep integration with industrial economy, the penetration of AI technology among different ethnicities, countries, regions, and populations differ to a great extent. The trend of widening and deepening digital divide will be more transparent. At the same time, under the manipulation of algorithm technology, information isolated islands and echo chambers make people more atomized. Those who enjoy the new conveniences brought by AI are uncertain about whether their humanity has ameliorated or not. Scientific culture will inevitably face the new paradox.

Fourthly, how will AI affect the progressiveness of scientific culture? Science and technology are crucial for propelling social productivity and improving people's well-being. Therefore, scientific culture should help promote social progress, and its progressiveness can be better reflected through civilized institutions. AI technology is constantly transforming people’s production modes and lifestyle, giving rise to various new governance challenges. For example, how to define knowledge and data sovereignty? Is it just and fair to treat special groups differently based on social scoring? How to prevent malicious manipulation of public opinion and political elections, or inducement of behavior among vulnerable groups? How to avoid the explosive expansion of the digital divide? Should machines have decision-making power? How to deal with accountability arising from the failures of AI applications? At present, there are numerous issues involving politics, legal systems, social justice, security and stability, ethical norms, and so on. It is the field where scientific culture will make significant contributions in the next phase.

In the past, the roaring of trains for the first time, the explosion of atomic bombs, the introduction of electric lights into homes, and the advent of the internet all made people stunned. Indeed, scientific and technological advancements have brought convenience, variety and comfort to human lives. It is widely agreed upon that the development of science and technology should be promoted or enhanced. I hold high expectations for AI technology, firmly believing that we can harness this technology, safeguard against and defuse economic, technological and social risks, and ensure that it brings greater benefits to humanity.

Source: Official Wechat Account of CAST