Shu Gao was educated in Nanjing University and the University of Southampton, UK, for M.Sc. and Ph.D. degrees, respectively. His research career includes the positions at the Second Institute of Oceanography (SOA), the Southampton Oceanography Centre (UK), the Institute of Oceanology (Chinese Academy of Sciences), Nanjing University, and East China Normal University.
His research interests include sediment transport and accumulation in shallow seas, evolution of coastal and shelf geomorphology, land-ocean interaction in the coastal zone, formation and evolution of Holocene sedimentary systems, water and sediment exchange in estuarine and coastal embayment, shallow marine material cycling, marine resources development and environmental management, and coastal engineering and environmental impact assessment. Since 1985 he has completed more than 400 publications in academic journal and book contributions, in the fields of marine sedimentary geology and environmental dynamics.
He works as a professor in marine geology at Nanjing University and a visiting professor of the East China Normal University, and is currently an Editor-in-Chief of Marine Geology (Elsevier) and the founding editor of Anthropocene Coasts (East China Normal University and Springer).
（1）What is the professional achievement you are most proud of?
It was a joyful experience to catch up with the era of reform and opening up and be able to enter Nanjing University for study. However, I wasn’t prepared for majoring in “Geomorphology and Quaternary Geology” initially. So, whether it was during my undergraduate or master’s studies, I couldn’t find a clear direction to strive for, which caused me a great deal of distress. Eventually, I decided to engage in research work on marine sedimentary dynamics after I arrived at the Second Institute of Oceanography in Hangzhou and settled down. However, I had no idea about the scientific problems involved, what achievements I would make in terms of scientific theories, new methods, and new discoveries in the future. Even when I went to study abroad in the United Kingdom, this question still troubled me greatly.
Although I devoted a lot of effort to research work for a long period of time and published a number of academic papers, the research outcomes do not seem to be satisfactory. The main focus of my efforts was on quantitatively characterizing the movement of sediments and the evolution of landforms that we could observe, but I couldn’t boast of any professional accomplishments to be proud of. Over the years, I proposed several research topics for writing academic papers. Some of the ideas that emerged and were implemented in my research since 1981 are as follows:
(1) Sandy sediment on tidal flats is transported from the sea to the shore due to tidal deformation mechanisms.
(2) According to the formula for mean vertical flow velocity, the thickness of muddy sediment deposition on tidal flats is controlled by the composition of sediment sources, while the range of deposition on mixed mud-sand flats is influenced by tidal processes.
(3) Digital image processing techniques can improve the method of grain size trend analysis.
(4) In addition to empirical observation and induction, the trend of grain size can also be simulated through dynamic processes to confirm its correlation with sediment transport.
(5) By applying the balance equation of natural tracers, the net transport direction of sediments can be determined.
(6) The steep slope at the edge of a tidal flat salt marsh may be a self-organizing phenomenon in depositional environments and should be given attention when extracting information on coastal erosion and sea-level rise.
(7) The P-A relationship in equilibrium tidal channels can be established not only through regression analysis but also through sediment dynamics. The exponent n in the power function expression is approximately 1.15.
(8) The effectiveness of inductive reasoning involves an interesting logical problem: it does not conform to the principle of the relationship between premises and conclusions. However, this is related to a long-standing misunderstanding of the process of inductive reasoning. From the perspective of mathematical logic, inductive reasoning can be expressed in a logical form. The key lies in whether the redefined inductive reasoning can be accepted by people. Perhaps traditional concepts can be referred to as “Type I Inductive Reasoning,” while the new perspective can be referred to as “Type II Inductive Reasoning.”
(9) Linear regression analysis is usually performed using the Gaussian method, but people often overlook the prerequisites of the Gaussian method when applying it. When these prerequisites are not met, the formula for linear regression analysis should be appropriately modified.
(10) Coastal zone development involves various aspects such as resources, environment, ecology, and disasters. Currently, the management in this field is lagging behind, and a “robust management model” should be developed, with computer technology providing the necessary data and processes for management.
(11) For source analysis, in order to achieve quantitative goals, the tracer markers should have a certain quantity, but for them to be effective, the tracer markers should possess stability, or the changes during transport should be known.
(12) Personal diaries containing detailed weather records can serve as historical materials for studying spatial patterns of climate change.
(13) According to the sediment balance equation, the landward part of a river delta has a growth limit. If the relevant parameters in the balance equation are known, the size of the delta at the point of reaching the growth limit can be determined.
(14) The preservation potential of sediment layers in tidal flat environments can be used to assess the first-order completeness of sediment sequences, which refers to the ratio of the preserved portion of short-term scale sedimentation to the total sedimentation that has occurred.
(15) Parameters of the boundary layer in extremely shallow water environments can explain various hydraulic and geomorphic phenomena on tidal flats, such as the advancing front of the rising tide, “channels within channels” in tidal channels, and the planar bedform of silt-fine sand flats after low tide.
(16) The distance from the aboveground stem of Spartina alterniflora to the point of bifurcation of the underground rhizome is relatively fixed. Therefore, by determining the position of buried plants, information on sedimentation rates can be obtained.
(17) Tunicates can attach and grow on Spartina alterniflora salt marshes on tidal flats. The variation in calcium carbonate content of sediments over time can be studied to evaluate the carbon sequestration effect of this factor.
(18) Sedimentary records contain many parameters with environmental significance, such as sedimentation rate and organic carbon content. When the range of variation within a certain period is small, it becomes a characteristic value. By simulating the dynamic process of its formation, the environmental characteristics can be revealed.
(19) For a delta complex, sedimentary records from different periods may be distributed in different locations, such as the shifting positions of delta lobes. Therefore, whether the sedimentary record is complete or not depends on the results of piecing together these fragments. The completeness shown after piecing together is referred to as the second-order completeness of sedimentary records. The Yangtze River sedimentary system is a good example.
(20) Long-term growth simulation of coral reefs can be conducted based on the mass balance equation. The dynamics of the Greenland Ice Sheet are also analyzed using a similar approach.
(21) In fluvial sedimentation, distal mud exhibits two forms of clinoforms. One is the aggradational clinoform that shrinks towards the sea due to subsidence, while the other is the progradational clinoform that advances parallel to the sea due to gravity flows. The situation is slightly different for subaqueous deltas at river mouths, but similar clinoforms will eventually form.
(22) The initiation conditions of sedimentation involve not only the classic five factors (particle size, particle shape, grain size distribution, bed shear stress, particle viscosity) but also consider the turbulent factors of water flow.
(23) Intensity information of storm deposits can be extracted using the “solution space contraction method.” This involves simulating the possible storm intensity ranges in sedimentary records from multiple locations and taking their intersection to narrow down the potential storm intensities to the lowest range.
(24) Changes in extreme events such as storms under the influence of climate change vary at different locations, so geographical zonality must be considered. It is not only a factor influencing spatial distribution patterns but also a factor affecting spatial distribution changes.
(25) Ocean science has two perspectives: viewing the ocean from land, which is known as British-style oceanography, and viewing the globe from the ocean, known as American-style oceanography. In the future, a new field of oceanography with a focus on human impact may emerge.
(26) By applying the sediment balance equation, the total amount of sedimentation in geological history and its influence on sedimentary records can be analyzed.
(27) The classification of river deltas should be based on factors such as runoff, tides, waves, shelf currents, initial topography, and sea level position, constructing a continuous spectrum. The triangular diagram method based on runoff, tides, and waves alone is insufficient.
(28) Quantitative expressions of mudcrack morphology can be applied to extract environmental information.
Among them, there are very few highly important issues, and most of them merely belong to the category of modifications and supplements to the original theory. Due to this reason, significant breakthroughs have been rare in terms of theorical innovation.
Relatively speaking, the first innovative progress lies in the analysis of sediment particle size trends. There is a wealth of data on sediment particle analysis, but what is the use of this data? For a long time, there has been a great paradox. Some scholars believe that although there is a large amount of particle size data, it does not explain the problem, and it cannot even be considered statistically significant. They argue that it does not have much significance in resolving geological problems. On the other hand, another group of scholars believes that since sediment particle size is related to the environment, it must contain information about material transportation, deposition, and geomorphic evolution. These are two completely different perspectives. On the positive side, I agree with the latter view. Data always requires interpretation, and there needs to be a theoretical explanation as to why sediment grain size data appears as it does. The connection between sediment grain size parameters and sediment transport patterns is a good breakthrough point. For this reason, I have made considerable efforts and published a series of papers in the early 1990s, presenting my own developed quantitative analysis methods. Interestingly, this group of papers is the most frequently cited among all my published works.
Another effort is attempting to explore sediment transport and its environmental effects through sediment budget analysis. This issue initially arose from the question of whether sediment in estuaries is transported inward or outward. When I was working on my doctoral thesis, I pondered over this for a long time. One night, I had a dream and found a way to answer this question, which was to establish a sediment budget equation. As soon as I woke up after the dream, I quickly recorded what I had thought in the dream. The next day, I carefully considered whether it was valid, and fortunately, the deductions I made were valid. Later, this research became a part of my doctoral thesis. The main results of the study were published in 1995, and it was the only paper among all my works that was published without any modification requested by the peer reviewers. Years later, when considering issues such as sediment carbon budget, tidal flat development, and delta growth, I realized that sediment budget was also crucial. For example, based on this budget model, I analyzed the growth pattern of the Yangtze River Delta and proposed the concept of growth limit for deltas. This concept cannot be induced from observational data; it is a new approach that connects theoretical assumptions with measured data through generalized simulations. In fact, it is a special case of a complex system. Now I am trying to summarize this method as “survey simulation” or “AI mimicry simulation” and attempting to apply it to the study of reef sedimentary systems formation and evolution. Now I believe that although this is only a special case of a complex system, this special case itself is worth finding many examples for. If one day many special cases of complex systems can be brought together, the general theory can gradually be improved. This task is too massive and far beyond individual capabilities, and in the future, it will rely on artificial intelligence. In fact, the design of artificial intelligence is based on network theory of complex systems.
（2）As a scientist, what do you enjoy most?
Regarding the pleasure of scientific research, Feynman’s views are profound. He said, there are three values of scientific research: scientific knowledge enables us to open a new door to enter the unknown world, research brings with it the fun called intellectual enjoyment, and science tells us that any scientific result has its uncertainties. Over the years, the members of my research group and myself have been practicing these principles and have deeply felt their impact. In the past, we followed our teachers to work on applied projects, or projects funded directly by non-governmental entities, using our knowledge of geomorphology to provide solutions to the various problems. At that time, the significance of Feynman’s ideas was not apparent to us. It was only after entering the field of basic research that we gained new insights.
From the perspective of basic research, its greatest driving force comes from researchers’ passion which is driven by two factors. The first is loyalty to science, a loyalty that borders on a religious relief in science. The second is interest in scientific problems. Indeed, scientific problems have many intriguing aspects. Earth sciences originates from natural history, where many natural phenomena are so bizarre that they require scientific explanations. However, traditional natural history alone cannot provide complete scientific explanations. It requires the support of knowledge in mathematics, physics, chemistry, and life sciences. How to integrate the knowledge of these disciplines into Earth sciences is a significant challenge. Yet there are also many exciting aspects to this endeavor. The more interdisciplinary the scientific problem being studied, the more it ignites the enthusiasm for research. The problems of Earth sciences involve a wide range of spatial and temporal scales, process mechanisms, and the relationship between theory and observation, offering endless possibilities for exploration.
At the same time, scientific research as a profession, coupled with the role of a university teacher, allows the alignment of personal interests with one’s occupation. In other words, the work one desires to do coincides with the means of livelihood. In this case, there is no reason to be lazy or not to work hard. It’s precisely because of this alignment that, over the years although the problems I researched may not be the most important in science and lack significant achievements, I have maintained high academic productivity, undertaken a considerable number of teaching assignments, and authored textbooks. I do not regret this at all.
（3）Which places around the world has your scientific journey taken you to, and what was your most memorable field experience?
Due to the nature of Earth science itself, we need to conduct fieldwork for data collection. In addition, there are many opportunities for field trips while participating in various academic activities, such as attending international conferences. As a result, I have been to many different places. In formal field work, some locations are relatively rare, such as participating in Atlantic and Mediterranean voyages, conducting fieldwork in the English Channel and the North Sea region, and going out to the Bohai, Yellow, East China and South China Seas. There are also places that I frequently visit, such as the coast of Jiangsu, the bays of Zhejiang, Hainan Island, and the coast of Shandong. Over the years, the number of days spent on fieldwork has accumulated to about five years.
The most memorable experience was the Integrated Ocean Drilling Program (IODP) Expedition 333 during December 2010 and January 2011. During the expedition, scientists on board worked in two shifts, with each shift working for 12 hours. The chief scientist led daily academic seminars, creating a highly scholarly atmosphere. There were great expectations for sample collection during each shift. The ship had a well-equipped library with a wealth of reference materials. Apart from work, the library served as the main source of information. While working, it was important not only to keep detailed work records but also to jot down the scientific question sand research topics for further investigation. By the end of the expedition, my work record exceeded 100,000 words. After returning, I organized the findings into a research paper and wrote over 50,000 words for a popular science article, which was later included in the book “Towards the Blue Coast”. From this expedition, I experienced many new methods of field data collection and theoretical analysis, expanding my research horizon. Although I was 55 when I participated in this event, it still had a significant impact on my personal scientific research. As long as one works hard, it is never too late.
(4) What are some of the most recent and exciting scientific advancements in your respective areas of research?
In scientific research, writing each paper brings a great sense of accomplishment because it involves a significant amount of effort and time investment. So far, I have written over 500 papers. Just imagine the satisfaction that comes with completing each one, it’s such a wonderful feeling. However, as philosophers have pointed out, this pleasure is short-lived and fades away shortly after publication. Thus, there is a need to write new papers in order to experience the same joy once again. Therefore, it can’t be said that there have been exciting breakthroughs, but it is indeed enjoyable. The objective measure of success lies in the feedback from peers. Any research must have an impact on fellow researchers; otherwise, the academic contribution is minimal.
Now our research has entered a new phase. The entire study has gone through three stages. The first stage focused on sedimentary dynamic processes, sediment transport, accumulation processes, and their impact on geomorphological environments, which were the core issues. The second stage, starting around 2003, concentrated on the process-product relationship, particularly the formation of sedimentary records and methods of extracting information from them. We are currently in the third stage, attempting to understand the position of sedimentary dynamic information within the entire Earth science system and contribute to solving integrated scientific problems. For instance, in the post-plate tectonics era, the role of sediments as participants and recorders in global sedimentary environmental evolution, the role and impact of sediments in the global ecosystem, and the influence of sediments on climate and environmental changes. The papers published in these three stages are different, yet they have all brought great joy.
（5）What qualities should a successful scientist possess?
What are the traits required of a successful scientist? This seems to be a difficult question to answer. There are many different criteria for success, and success requires a variety of conditions. However, we may refer to the views of predecessors on this issue.
Einstein believed that there are three types of scientific workers. The first type sees scientific research as their profession and livelihood. The second type has a strong interest in the scientific problems they study. The third type is loyal to science and has scientific faith to support them. He said that only the third type of person can achieve the greatest progress in science. Having scientific faith enables a person to maintain their pursuit of science even under difficult conditions, which can be considered a trait.
Another point is persistent effort. On this aspect, we need to learn from Darwin. He mentioned having strong powers of observation, but slightly weaker theoretical abilities. Therefore, he emphasized playing to one’s strengths, observing more, recording more, and analyzing more data. Through persistent and unwavering effort, new ideas can eventually emerge. Therefore, traits such as improving observational abilities, developing data analysis skills, engaging in scientific thinking without fearing difficulties, and maintaining persistence are important for pursuing excellence in the field of knowledge.
Scientific faith and perseverance are the traits of outstanding scientists. However, these traits only come into play under specific conditions and do not guarantee the emergence of great achievements. These excellent traits are important but not sufficient conditions for success. They are merely necessary conditions.
（6）If you were not a scientist, what other profession would you choose to pursue?
Scientific research is a very rewarding endeavor. However, for me personally, entering the field of scientific research was to some extent determined by the college entrance examination. After the exam, I was not admitted to the major of my choice but was instead assigned to a major in Earth sciences. For the candidates, the name of this major was difficult to understand, and as many researchers in this field concluded that it was difficult to enter this profession without long-term nurturing and guidance from the predecessors. So, at that time what one would study and what profession they would pursue was to a certain extent a matter of fate, rather than one’s own choice. Nevertheless, fortunately, I eventually developed a passion for Earth sciences and worked diligently in the field for many years.
When I was choosing my major for the college entrance exam that year, I was most interested in the fields of literature, history, and philosophy, hoping to be admitted to a humanities program. I scored well in Chinese and English exams, which to some extent reflected my efforts in that direction. However, I ultimately gave up and took science instead. The primary factor behind my initial choice of humanities was perhaps influenced by my grandfather. I was interested in writing and translation and wanted to make a living as a humanistic scholar. Sometimes, even now, I still feel the urge for such a desire. So, if I had chosen humanities and been admitted back then, perhaps I would be engaged in writing and translating in the fields of literature, history, and philosophy now.
When I was in high school, a group of my classmates and I were also very interested in traveling. We often went out together and climbed every peak around our town. Once, we even reached the highest peak visible before us, at an altitude of about 1,100 meters under the guidance of our PE teacher. It was already early summer at the foot of the mountain, while there was still residual snow at the summit. After high school, I worked as a lathe operator in a factory, but I still often looked for opportunities to take short trips and climbed the highest mountain near the factory. If combined with humanities, I used to think that being a journalist would be a great profession. It would provide frequent opportunities for travelling, encountering interesting things, and writing about and reporting on them.
（7）What activities do you engage in outside of work?
Doing scientific research is extremely busy. Therefore, it is very difficult to arrange other activities outside of work. According to the sociologist Robert K. Merton, there are four responsibilities or scientists: scientific research, student training, management of research activities, and social service. Scientific research is the primary job, requiring one to find their position within their field, conduct data collection and analysis, build computational models, and develop theories and methods. The workload is substantial. In this process, it is necessary to establish a research team, including postdoctoral fellows, doctoral and master’s students, and laboratory technicians. Therefore, while conducting research, one also needs to cultivate and train this team. Graduate students need to get their degrees and postdoctoral fellows need to find jobs. In this regard, there is a certain responsibility towards them. Daily management of research activities and research institution management require time investment to plan academic development, create and provide adequate working conditions so that the personnel in the institution can be competent in their research. Organizing and coordinating daily affairs, as well as having contingency plans in case of crisis, are also necessary. Social service mainly includes participating in activities of academic organizations, editing and reviewing academic journals, and providing social consultations. With all these responsibilities combined, the workday is insufficient and even weekends are occupied. Academic paper writing, ironically, needs to be squeezed into whatever spare time is available. Due to the above reasons, opportunities to engage in other activities are rare. When I was young, I had many hobbies, such as stamp collecting, fishing, playing chess, bridge, collecting, and playing music. Later, under the pressure of research work, I eventually gave up all these hobbies.
Many researchers have had similar experiences. While reading Darwin’s Recollections of the Development of my mind & character, I noticed that Darwin once mentioned his interest in music but had to give up that hobby for his research work. He expressed a wish that if he could go back in time, he hoped to continue these hobbies. The feeling of both having hobbies and reluctantly giving them up is something that I believe many scientists share.
However, some hobbies can be combined with work. For example, I have a keen interest in travelling, and the nature of our work often requires us to go on field trips, sometimes even abroad. In such cases, one can make full use of the opportunities to travel, appreciate the architecture, museums, and social life of a city, and document them through photography. This can to some extent satisfy the hobby. Among all the scientific fields, I believe that if one engages in research in physics or chemistry, it may be even more challenging to maintain hobbies. In the field of Earth sciences, at least there are some opportunities to travel and directly observe nature, cities, and human society.
（8）How do you balance work and personal life?
To be honest, I am poor in balancing work and life. It is difficult to balance work and hobbies, but it is even more difficult to balance work with life. What constitutes a good life? This question itself is very confusing. Some aspects can be determined, such as a good life should bring happiness to one’s family and make each day go smoothly. This requires financial stability and also demands sufficient time investment. The two are also interrelated. If one’s financial strength is weak, investing more time is not feasible. Conversely, even with excellent financial conditions, if little time is devoted, it becomes meaningless. For scientific researchers, this can be a major problem. In reality, many renowned scientists tend to sacrifice other aspects of life due to their dedication to work.
Therefore, if there is an outstanding talent within a family, other family members are bound to make considerable sacrifices. In the Encyclopedia Britannica, the names of 24 members of the Bernoulli family appear, including mathematicians, physicists, chemists, and engineers. One can imagine the price paid by other family members for the achievements of these individuals. For those of us who have not achieved the status of great scientists or scholars, we feel deeply guilty for the sacrifices made by our families. The question is, what can be done to avoid this guilt? Unfortunately, there may be no solution. Because before achieving great accomplishments, we do not know whether we will achieve them or not. The inability to achieve significant accomplishments is an answer that comes after years of effort, and during this process, the sacrifices of family members have long been made.
This situation also illustrates the importance of research management. If we can improve the efficiency of research management, allowing researchers to work no more than ten hours a day and no more than six days a week, they would have some time to spend with their families and lead a good family life. It would also provide them with some time to invite friends for gatherings, where relaxed and enjoyable atmospheres may even facilitate better academic ideas and lead to scientific achievements.
(9) During the COVID-19 period, you conducted research on shepherd’s purse and also shared the joy of reading “Weeds: In Defense of Nature’s Most Unloved Plants.” Could you please tell us more about it?
The COVID-19 pandemic was a bad thing as it significantly limited our work and life. However, there are exceptions. In early March 2020, when I returned to work from home, I was required to quarantine for two weeks, and even after that, I couldn’t go to the office for a long time. During this period of isolation, my interest in botany suddenly sparked. Previously, while browsing in a bookstore, I came across the book “Weeds: In Defense of Nature’s Most Unloved Plants.” The author described 120 species of weeds in his garden and the surrounding environment, writing an essay for each species. The 120 beautifully written essays, rich in knowledge, provided a delightful reading experience. The book mentions shepherd’s purse, a wild plant that people consider delicious. In real life, there are many food products made from shepherd’s purse in supermarkets in Shanghai, such as shepherd’s purse dumplings, buns and wontons. It is evident that people love this delicacy.
In early March 2023, the shepherd’s purse flowers were blooming in abundance just outside my dormitory, covering the area in white. So, I thought, during these days of quarantine, why not conduct a formal field investigation and collect valuable data? Previously, when conducting vegetation surveys on tidal flats, we collected various basic information about plants within the “quadrats,” such as the number of plants, aboveground biomass, and plant morphology. Therefore, I used the same method to collect data. I selected three 2x2m quadrats in the young faculty apartment complex, representing environments with vigorous, moderate, and poor growth conditions. I found that the growth of shepherd’s purse was closely related to the soil and moisture conditions of its location. It thrived in moist areas but fared poorly in dry areas. The growth of the plants was directly reflected in their morphology. During the process of collecting shepherd’s purse plants, I categorized them based on their morphology and discovered four basic types. The proportions of these different plant types varied significantly among the three quadrats, indicating the environmental quality. Interestingly, this aspect was not mentioned in the literature on shepherd’s purse. There were also vegetable plots in the complex, where shepherd’s purse grew fastest because the soil and moisture conditions were excellent. This confirmed the relationship between plant morphology and the growth environment.
Next, I collected data from five aspects: roots, stems, leaves, flowers, and fruits. For each plant type, I obtained data in these five categories and made comparisons. After obtaining the basic data, I wrote a draft paper titled “Shepherd’s Purse Plant Morphology and Survival Strategies.” Although I didn’t have the intention to submit the draft paper, I felt that it would be meaningful if it could be transformed into a popular science article. Such an article should have the desired impact. However, the existing data lacked a highlight. To attract readers, a popular science work must have an important knowledge point that builds up to a climax. For several weeks, I couldn’t find that point.
Working at the university began to return to normal in May. One day, I went to the library at Zhongbei Campus to attend a meeting. In a pile of returned books, I found a book on applied mathematics that mentioned the Fibonacci sequence, the golden ratio, and plant morphology based on the Fibonacci sequence. This gave me a crucial inspiration: shepherd’s purse morphology could also be described using this method. Shepherd’s purse plants are small and occupy a tiny space, but within this space, they must fully utilize solar energy to produce more seeds. In my previous morphological studies, I discovered that the angle between every two siliques on the inflorescence of shepherd’s purse was approximately 137.1 degrees. I also observed other cruciferous plants, such as rapeseed grown in the garden, and found that this angle was close to that number as well. Is there any connection between this angle and the Fibonacci sequence? I reexamined the angles of dozens of shepherd’s purse plants, and their average value was close to 137.5 degrees. The most significant correlation was (360-137.5)/360 = 0.618, which represents the golden ratio. This angle allows for the most efficient utilization of solar energy and the production of the maximum number of seeds. This was the very highlight I was seeking. Soon after, I wrote the popular science article “Survival Strategies of Shepherd’s Purse” and published it in the journal “Science.” The arrangement of shepherd’s purse siliques is most favorable for its survival and reproduction, and all of its survival strategies aim to produce more seeds. This amateur research fulfilled my interest in botany to a considerable extent and brought a sense of accomplishment during the days of COVID-19 isolation.
(10) Your recent research on the Triassic carbonate sedimentary dynamic processes follows the approach of modelling AI. How do you think scientists can better utilize AI for research in their own fields?
Previously I published a paper titled “Process-product relationships of atoll deposition systems: a preliminary testing of exploratory modeling” in Oceanologia et Limnologia Sinica. The aim was to study traditional sediment dynamics problems from the perspective of AI methods applied to complex systems. The research on Triassic carbonate sedimentary dynamic processes follows the same approach, aiming to demonstrate the applicability of this method regardless of the era or environmental conditions.
AI as a research tool has developed faster than we anticipated. In the near future, AI-generated writing will replace researchers’ paper writing. Therefore, the way researchers write papers will undergo significant changes, and it is important to adapt early. There is also an urgent question of how researchers can harness the potential of AI to write more and better papers.
Traditionally, our scientific training involved paper writing based on field data collection, laboratory analysis, data processing, and interpretation. Alternatively, modelling was conducted by constructing initial and boundary conditions based on the collected data and then performing comparisons and validation.
However, this approach to paper writing is unsustainable in the long run. It relies heavily on data collection and inductive reasoning. Mathematical modelling depends on certain physical principles to construct governing equations, but solving these equations requires computational power. Moreover, as research progresses, it becomes increasingly difficult to find groundbreaking topics. While our research may be new in the sense of “digging deeper into unexplored soil like earthworms,” its significance diminishes. This is primarily because previous researchers have done extensive work in terms of theory, methods, and techniques, making substantial breakthroughs difficult to achieve. I have previously written a blog post titled “Analysis of Existing and Potential Earth Sciences Papers” on the website Sciencenet, attempting to describe the characteristics of current Earth sciences papers and estimate the number of papers that need to be written. It is indeed challenging for humans to complete all these papers using traditional methods.
However, the approach of artificial intelligence is different. It does not start by searching for answers through data collection but rather integrates existing knowledge and obtains insights through macroscopic system analysis. The methodology involves constructing a system in which all variables have defined domains, allowing characterization using the features of complex systems and general equations of evolution. The continuous and momentum equations in hydrodynamics are just specific examples of this methodology. AI analysis is not specific to particular environments; it encompasses all environments. It links influencing factors and variables to the behavior and evolution of systems and compares them with critical judgments in the real world. When the judgments align, it can infer the fundamental characteristics of the system and construct specific data highly similar to the actual system. This is quite different from our approach, which requires data to be obtained through field observations and sampling. AI does not need to follow this process; once a sufficient amount of observational data is available, it is enough. In terms of efficiency, our data and research methods are inefficient. The past is in the past, and while the burden of data collection continues to increase, our understanding of the system has not significantly improved. AI, on the other hand, is different. It may have more errors initially, but it has a promising future and can continuously progress, ensuring sustainability.
At our current stage, the approach of modelling AI is intriguing. The technical aspects of AI itself are not our concern. Professionals in this field will push it forward on a large scale. The key point is for us to consider how AI would write papers if it were to do so. By modelling its writing approach, we can examine the resulting papers and determine their prospects. If the results are promising, AI-generated papers will soon become mainstream, and we should be prepared for that. When a Go master plays the current 19-line Go, a high degree of agreement with the moves suggested by AI would bring a significant sense of accomplishment. Similarly, if the articles we write align with future AI-generated papers, it would also provide a sense of achievement, wouldn’t it?