Computer science is one of the fastest growing disciplines. “Today, we carry more computing power on our smartphones than was available in early models that took up entire rooms” (Zimmermann, 2017). Since the birth of the first computer, human society has officially entered the era of computer technology revolution. With the efforts of computer scientists, computer science and technology have developed rapidly and penetrated people’s production and life at an extremely fast speed. Industrial automation, intelligent life, etc. have become the fruitful results of the development of computer science and technology. At present, the application of computer science and technology in various fields such as medical, education, military, etc. is becoming more and more extensive, and its application range is still expanding. At the same time, computer science and technology have also extended to new industries.

These achievements make people wonder: What is the limit of computer science? What will the future of computer science be? The purpose of this critical research essay is to answer these questions by exploring the unsolved computer science problems from a historical, scientific, and ethical perspective. It is no exaggeration that the solution to these problems will define the future of computer science and human civilization.

In addressing the future of computer science, in relation to the unsolved problems in computer science, it is critical to address historical, scientific, and ethical perspectives. First, a historical perspective provides the context of software applications. The earliest wave of computer application were simulations of the real world that began in 1960. Before 1964, computers were mainly used for scientific, mathematics, and military purposes due to the bulky sizes (Zimmermann, 2017). With the development in hardware, computers were increasingly applied in the business and individual scenarios, such as generating payrolls for major corporations, playing video games, and the recent wave of virtual reality. The ultimate simulation is the construction of an entirely virtual world as imagined by pop culture.

Another wave of computer application is in communication and storage, which started in the 1980s. E-mails, online booking, ecommerce, eBooks, social networks, etc. have been created to accommodate the communication needs of people. Similar to simulation, communication is also one of the major influences of computer science on human society. Since the 2010s, a new trend of artificial intelligence has been created, including speech recognition, smart home appliances, and machine learning. Among the three waves of computer applications, artificial intelligence is found to be the newest, least develop, and the one with high potentials.

Additionally, hardware technology has also gone through tremendous evolution in the past decades. The first computers were mainly made of electronic tubes. It used to be bulky and extremely inconvenient to carry around. The main purposes were for military, scientific computing, and experimental development (Waldrop, 2016). The current generation of computers can concentrate the operating and control functions on a single chip, which makes computers smaller, cheaper, and more portable. Past generations of computers point out that superior performance and increased portability has always been the main trend of development in computer hardware.

Since 1970, as predicted by Moore’s Law, the number of components that can be accommodated on an integrated circuit (microprocessor) is doubled every 18 months, and the performance of the computer is nearly doubled. However, this trend began to slow down after 2005, and extremely small integrated circuits are facing problems such as heat dissipation (Waldrop, 2016). If we continue to develop according to this trend, when the size of the integrated circuit is close to the atomic level, the motion of the electrons will no longer follow the classical laws of physics. This poses a serious challenge and limitation for computer science in the future.

Secondly, a scientific perspective provides an insight into the nature of computing, and the future of it. One of the defining features of computers is that they perform complicated tasks with fast speed. One unsolved problem of computer science, the P versus NP problem, will determine how fast the future computer completes its tasks. The problem states that if P represents a category of problems with efficient solutions, and NP represents a category of problems with efficiently verifiable solutions, then P=NP would mean that “for every problem that has an efficiently verifiable solution, we can find that solution efficiently as well” (Fortnow, 2009).

Many Turing Award winners agree that the P versus NP problem is a very important issue to be solved. This question was ranked first by the Clay School of Mathematics as the “Millennium Award Question” for the $7 million prize. Steve Smale, winner of the Fields Award, also listed the problem as one of the most important mathematical and computational problems for the next century. Steve Cook predicted that even if P=NP is false, it will take human beings approximately one century to find proof (Cook, 2003). This mathematical problem is of fundamental importance to computer science because many of the practical problems in real life have natural and non-deterministic algorithms.

Also, the operation of the computer is deterministic. Therefore, understanding the relationship between determined and non-determined times is particularly important for the next stage of computer evolution. The idea of combining quantum mechanics with computational problems thus enters the discussion. This idea was proposed by Feynman in 1982. According to his vision, a standard quantum system (easy-to-manipulate system) can be used to simulate complex quantum systems, thus solving unsolvable quantum problems for the classical computer, especially the problems related to quantum multi-body physics (Boghosian, & Taylor, 1998). Although people did not know how to implement such a quantum simulator at the time, Feynman’s thought directly affected the development of quantum computing.

P. W. Shor examines why there are few quantum algorithm classes discovered so far, and he gives two explanations: First, quantum computers work differently from traditional computers, and the existing technologies are no longer useful for designing algorithms. Intuition is no longer useful for understanding the computational process. Second, the number of problems that quantum computers can provide substantial acceleration is relatively small compared to using traditional computers (Shor, 2003). Without solving these problems, the hardware capabilities of modern computers are stretching thin under the Moore’s law.

Finally, there is the ethical perspective of unsolved computer science problems to consider. Although artificial intelligence has been a popular computer science topic in recent years, the real artificial intelligence problem is yet to be solved. Till now, computers still largely fail to “understand.” Behind the seemingly intelligent interactions there is always human design. Lampson explains the two practical problems for computers to understand. The main challenges of the first problem are real-time vision, road models, vehicle models, and external object models that invade the road.

All this knowledge requires a driver to learn for many years (Lampson, 2003). The input of the sensor, the uncertainties in the operation of the vehicle, and the changes that may occur at any time in the environment can only be addressed by a true artificial intelligence supported autopilot. However, building computers that can “understand” may lead to ethical dilemmas such as the breach of personal privacy. To provide the best services means a total invasion into people’s lives, which may raise serious concerns in the future.

In addition to privacy, artificial intelligence often leads to greater concerns over the relationship between humans and machines. Automated programming is a specification in computer science that remains mostly unsolved. People have struggled for this problem for 40 years, but progress has been limited. In some areas, the description of the program design is feasible. Spreadsheets and SQL queries are successful: their specifications are close to programs. Yet still, these solutions use the spreadsheet macros, SQL updates, and precise control of the planning in HTML (Lampson, 2003). Without these controls, it is still impossible for the current level of artificial intelligence to perform complicated programming tasks.

However, once computers are able to program on their own, they will become more capable than human beings in most disciplines. This may lead to unemployment, poverty, discrimination, and other social problems. Once artificial intelligence grows too powerful, vulnerabilities in the Internet and vulnerabilities in the artificial intelligence technology itself can create huge security risks.

In analyzing unsolved computer science problems and their implications, the constructs of molecular informatics and AI ethics are important to consider, in order to gain a deeper understanding.  First, molecular informatics is defined as computing processes based on the construction of molecular structures (Ghose, 2017). Molecular informatics is fundamentally different from the current computer science based on 1s and 0s. Instead, it offers a three-dimensional alternative to the current mainstream computing.

Although the program is still in the initial stage of development, it reveals that humans should not be confined in the past achievements and should explore things in different directions to find a breakthrough of computing. Similar to the concept of quantum computing, attempts to revolutionize the boundary of computer science and actively integrate computer science with physics, chemistry, and biography may lead to the solutions to some of the unsolved computer science problems mentioned above, such as the P=NP problem and the creation of a true artificial intelligence.

Also, critical to gain deeper understanding is AI ethics. AI ethics is defined as the ethical rules that contribute to the responsible development of artificial intelligence (Castro, 2019). Large tech companies in the world such as Google and Microsoft have all established their own AI ethics guidelines. However, these efforts have not received the recognition of the public because the AI ethics drafted by large corporations have almost no legal binding power. Instead, they are more like publicity strategies that help these companies avoid government intervention.

In order to change this situation and create truly governing AI ethics, the government has to be involved in the process. For example, an AI ethical conduct alliance can be established and led by the government, whose members include the major tech companies in the country. Regarding the issues of unemployment, privacy, and security discussed above, members of the alliance must have absolute transparency on the issue. Violation of any establish AI ethical rules leads to fines and other punishment.

Therefore, I conclude that unsolved computer science problems are mainly in two aspects: the nature of computer and artificial intelligence. First, the nature of computing and the limitations of it. Second, the creation of true artificial intelligence and its implications on the human society. The first aspect of the problems is based in the history of computing, as a way invented by human beings to perform complicated tasks with higher efficiency. However, with the limitation in hardware, it is becoming increasing difficult to see a breakthrough in computing power.

Possible solutions to this problem lies in the future development of quantum computing, molecular informatics, etc., disciplines which aim to revolutionize the existing computing technologies. The second aspect of the problems is relatively new, because humans did not have the excessive computing power to develop artificial intelligence in the past. The unsolved problems in AI revolve around getting computers to “understand” and “program” without human intervention. In addition to the technical difficulties, these problems also lead to ethical dilemmas that need to be resolved.

This conclusion is significant because it sheds light on the important directions for the future computer science. The solution of any one of the above-mentioned problems will have explosive effects on the human society, significantly promoting the computation and production capabilities of human beings. Computer science and technology is one of the fastest and fastest growing technologies in the world. This research paper summarizes the status quo of development before exploring future prospects that are beneficial to the development of computer science and technology.

In the past decades, most of the advancements in computer science have been bound by the Moore’s Law, which is stretching thin. The future of computer science thus lies in the breakthrough on computational theory, quantum algorithms, molecular informatics, as well as true artificial intelligence. It should also be noted that none of these unsolved problems are isolated from the rest. Their relationship is quite inseparable, like the relationship between hardware and software. Therefore, the solution to any one of these problems will naturally lead to advancements in the others.