Challenges of Artificial Intelligence for economies and societies

Ph.D. Candidate, School of Business and Technology Management, KAIST

Anuujin Sanjaajamts

sanuujin@kaist.ac.kr

Introduction

The main goal of the field of artificial intelligence (AI), the creation of an intelligent machine that resembles human intelligence, is one of the most ambitious scientific goals ever. It is expected that general-purpose artificial intelligence (AI) technologies will bring significant differences in the socioeconomic environment¹. Unsurprisingly, AI applications developed in research institutes, universities, and companies moved into everyday life². The McKinsey report on the economic impact of AI shows that 70 percent of private industries are adopting at least one of the AI technologies by 2030, which can bring an economic output of around $13 trillion by 2030 to the world economy³. As AI demonstrates an increased significance in most of society, there exists AI’s potential to replace human labor at lower cost, potentially raise treatment costs at the hospital, increase economic inequality across countries, and performance gap between first runners on one side and slow adopters on the other side⁴. In general, AI can overcome humans’ intellectual limitations and creates new domains within social industries such as education, health, marketing, finance, and manufacturing. The AI experts anticipate that AI systems will likely reach human intelligence by 2075, and some scholars predict that further development of AI towards super intelligence may negatively affect humanity⁵. Hence, AI adoption has many benefits across the global economy in the context of increased efficiency and improved productivity. In addition, AI has benefits in developed economies where new higher-skilled jobs are likely to be created. Therefore, it is crucial to identify and understand the challenges of AI to decrease the dark side and increase its benefit in the socioeconomic environment. 

What is AI? 

The crucial concept of understanding the term ‘intelligence’ brings us to understand what AI is⁶. How do people understand things before making a decision? The Collins English Gem Dictionary defines intelligence as the “quickness of understanding” and deciding instead of doing things by instinct or automatically.  Fundamental knowledge of AI was to improve the everyday life of administrative operations and decision-making processes. However, the literature has offered various definitions of AI, each of which indicates any device that analyses its environment and makes the decision to maximize its chance to achieve a specific goal⁷. Since the birth of the field, AI research has had two different but interconnected goals. First is narrow AI, to develop a system that makes decisions on a specific task, and second is general AI, to design a system that can reach flexibility and adaptability as human intelligence⁸. However, all of the applications that have been developed until now are examples of narrow AI technology. Recent approaches are to make AI systems more general by learning from multiple tasks to solve problems.  

Challenges of AI

The implementation of AI technologies brings significant challenges to civil societies and economies (see Figure 1). Therefore, adopting policies and regulations that will be more responsive to the rapidly advancing pace of technology development will be necessary to address the risk of AI applications⁹. 

As AI becomes more prevalent, it will likely challenge cultural norms and act as an indirect barrier within specific sectors of society. For example, the possibility of imitating human labor at a lower cost has the potential to cause job losses and destruction due to technology¹⁰. Also, healthcare professionals must be prepared to change and adopt new technologies¹¹, which will bring change in the interaction with patients. The introduction of AI technologies will demand a significant amount of investment in the health industry, increasing the cost of treatment for patients¹². Therefore, AI could have a massive impact on the productivity growth of specific socio-economic sectors. Over time, innovation may result in more unequal slices of the pie if policymakers underreact to the coming changes¹³. 

Data challenge

The use of artificial intelligence to perform such tasks as making a decision, learning, and solving problems require the use of complex and ambiguous data, which is not easy to interpret¹⁴. Data can be used to discover hidden patterns, which are used to solve ever-increasing types of problems using machine learning techniques¹⁵. Creating systems significantly more advanced than the ones accessible currently, demands teaching them to understand causal relationships within big data, which requires transparency, reproducibility, and specific frameworks to share across industries, organizations, and governments¹⁶. The authors highlighted the complexities of using artificial neural networks in the interpretation of imagery and the dimensionality obstacle. The challenges associated with AI and big data integration have been discussed in several recent studies.

Technological and technology implementation challenge

One of the most compressing dangers of AI is techno-solutionism; when AI is merely a tool, it cannot be viewed as an elixir. There is an aura of neutrality and impartiality associated with AI decision-making in some angles of the public consciousness, resulting in systems being accepted as objective even though they may result from biased historical pronouncements or even obvious discrimination¹⁷. It has emerged that digital transformation is often the context for AI projects combined with the Internet of Things (IoT), Big Data, such as call center transformation using advanced analytics and AI or transforming operations using the IoT, advanced analytics, and AI. 

Political, legal, and policy challenges

Legal regulation requires knowledge about what will be regulated¹⁸. In the future, AI will likely be used in areas that require stability, clear goals, measurement, and long-term vision, such as highly automated public and private banks, as well as robotic diagnostics¹⁹. The introduction of automated cars, or self-driving cars, has been one of the hottest areas of robotics deployment, as they have repercussions across the entire automotive industry²⁰. Also, city planning as well, and image processing technology is used in everything from video-conferencing backgrounds to deep fakes that look photo-realistic²¹. In the transition from the Industrial era to the fourth industrial revolution, the related problems changed; thus, legal and policy principles will not be distinctive, but modernization, adjustment, and reformulation are necessary^{22, 23}. 

The organizational and managerial challenge 

Research has found that businesses face significant challenges regarding the implications of AI if they do not formulate a comprehensive strategy²⁴. AI adoption is likely to be evidence-based, which depends on the ease of use, financial returns, and trust that will determine its success. As an organization, an enterprise-wide security strategy refers to defining and implementing a strategy across the whole firm; thus, the agile firm needs to respond to the customers’ needs, adapt production delivery on-demand, and implement decisions according to the market generation²⁵, which requires the managerial capability of finding a way to redesign the process while accepting the new advanced technologies such as AI. Therefore, reflecting digital skills, the main challenge for all business industries is the lack of skilled staff and knowledge in emerging technologies, which would cause implementation challenges simultaneously. In practice, this implies that national governments should prepare themselves by training their workforce, developing their research potential, maintaining the competitiveness of their economy, and ensuring the ethical use of technology²⁶. 

The ethical challenge

The ethics and artificial intelligence issues go back to the early days of digital computing²⁷. Given the increasing influence of AI solutions on all aspects of our lives, we need to ensure that their development promotes human fulfillment. Doing so requires tackling ethical challenges such as bias, discrimination, denial of individual autonomy, resource, and rights, unexplainable outcomes, and actively promoting responsible decisions and AI implementation through a robust regulatory^{28, 29, 30}. 

Despite the all-challenges new business models, frameworks, rules, and policies for adopting emerging technologies are beginning to be designed, particularly for the benefit of society. Therefore, an infinite number of new ideas, methods, and processes must be reorganized and modernized for the benefit of people and organizations³¹. Moreover, it is becoming progressively clear that humans and AI need to work together to create an AI assistant that can outperform either alone³².  

ConclusionArtificial intelligence absolutely can change human lives for the better. It is happening now in our everyday life. Although advances in AI have been unpredictable, significant progress has been made since the establishment of the field sixty years ago. Businesses, education, medicine, and governments worldwide have grown to see that AI technology is not only integral to their business model but also an integral component of their plans. The vital question that we need to ask ourselves is whether we expect artificial intelligence to be socially responsible or whether we expect it to be satisfactory for individual use only. The dilemma is whether the automatic car saves the driver from the accident or saves others on the road, but who would like to buy a vehicle that needs to be protecting others on the road rather than the driver? Numerous similar situations pose a social dilemma, and the answer to whether we want artificial intelligence prosocial or not certainly has no universally valuable answers³³. Therefore, developers, practitioners, and policymakers at all levels need to understand the challenges of artificial intelligence technology. Recognizing the potential challenges of artificial intelligence helps us to avoid the dark sides of this technology and embrace the positive prospective of its technological development for better socio-economic development.

읽을 거리

Stuart Russel & Peter Norvig, Artificial Intelligence: A Modern Approach. Third Edition, Pearson Education, Inc., 2010, 2003, 1995.

The main unifying theme is the idea of an intelligent agent. The primary goal of this book is to deliver the ideas, which have emerged over the past fifty years of artificial intelligence research and the past two millennia of related work including fundamental ideas of artificial intelligence.

Susan Leigh Star (1989). “The Structure of Ill-Structured Solutions: Boundary Object and Heterogeneous Distributed Problem Solving”, Distributed Artificial Intelligence, Pages 37-54. DOI: https://doi.org/10.1016/B978-1-55860-092-8.50006-X.

Research Notes in Artificial Intelligence: Distributed Artificial Intelligence, Volume II focuses on the growing interest in Distributed Artificial Intelligence (DAI). The selection offers information on a unified theory of communication and social structure and boundary objects and heterogeneous distributed problem-solving. 

---

[1] Fatima, S., Desouza, K. C., & Dawson, G. S. (2020), “National strategic artificial intelligence plans: A multi-dimensional analysis.”, Economic Analysis and Policy, Vol. 67, pp. 178-194.

[2] Littman, M. L., et al. (2022), “Gathering strength, gathering storms: The one hundred year study on artificial intelligence (AI100) 2021 study panel report.”, arXiv preprint arXiv:2210.15767.

[3] Bughin, J., et al. (2018), “Notes from the AI frontier: Modeling the impact of AI on the world economy.”, McKinsey Global Institute.

[4] Littman, M. L., et al. (2022), “Gathering strength, gathering storms: The one hundred year study on artificial intelligence (AI100) 2021 study panel report.”, arXiv preprint arXiv:2210.15767.

[5] Müller, V. C., & Bostrom, N. (2014), “Future progress in artificial intelligence: A poll among experts.”, AI Matters, Vol. 1, No. 1, pp. 9-11.

[6] Valle-Cruz, D., et al. (2019), “A review of artificial intelligence in government and its potential from a public policy perspective.”, Proceedings of the 20th Annual International Conference on Digital Government Research, pp. 91-99.

[7] Russel S. & Norvig P. (2010), Artificial intelligence—a modern approach 3rd Edition (3rd ed.), New Jersey: Pearson Education, Inc. 

[8] Littman, M. L., et al. (2022), “Gathering strength, gathering storms: The one hundred year study on artificial intelligence (AI100) 2021 study panel report.”, arXiv preprint arXiv:2210.15767.

[9] Ibid.

[10] Hunt, W., Sarkar, S., & Warhurst, C. (2022), “Measuring the impact of AI on jobs at the organization level: Lessons from a survey of UK business leaders.”, Research Policy, Vol. 51, No. 2, 104425.

[11] Wiljer, D., & Hakim, Z. (2019), “Developing an artificial intelligence–enabled health care practice: rewiring health care professions for better care.”, Journal of medical imaging and radiation sciences, Vol. 50, No. 4, pp. S8-S14.

[12] Sun, T. Q., & Medaglia, R. (2019), “Mapping the challenges of Artificial Intelligence in the public sector: Evidence from public healthcare.”, Government Information Quarterly, Vol. 36, No. 2, pp. 368-383.

[13] Littman, M. L., et al. (2022), “Gathering strength, gathering storms: The one hundred year study on artificial intelligence (AI100) 2021 study panel report.”, arXiv preprint arXiv:2210.15767.

[14] Hydén, H. (2020), “AI, norms, big data, and the law.”, Asian Journal of Law and Society, Vol. 7, No. 3, pp. 409-436.

[15] Littman, M. L., et al. (2022), “Gathering strength, gathering storms: The one hundred year study on artificial intelligence (AI100) 2021 study panel report.”, arXiv preprint arXiv:2210.15767.

[16] Ibid.

[17] Ibid

[18] Hydén, H. (2020), “AI, norms, big data, and the law.”, Asian Journal of Law and Society, Vol. 7, No. 3, pp. 409-436.

[19] Ibid

[20]  Littman, M. L., et al. (2022), “Gathering strength, gathering storms: The one hundred year study on artificial intelligence (AI100) 2021 study panel report.”, arXiv preprint arXiv:2210.15767.

[21] Malik, R. (2021), “AI augmentation in the field of digital image processing.”, International Journal of Computer Applications in Technology, Vol. 65, No. 3, pp. 235-244.

[22] Hydén, H. (2020), “AI, norms, big data, and the law.”, Asian Journal of Law and Society, Vol. 7, No. 3, pp. 409-436.

[23] Littman, M. L., et al. (2022), “Gathering strength, gathering storms: The one hundred year study on artificial intelligence (AI100) 2021 study panel report.”, arXiv preprint arXiv:2210.15767.

[24] Sun, T. Q., & Medaglia, R. (2019), “Mapping the challenges of Artificial Intelligence in the public sector: Evidence from public healthcare.”, Government Information Quarterly, Vol. 36, No. 2, pp. 368-383.

[25] Brock, J. K. U., & Von Wangenheim, F. (2019), “Demystifying AI: What digital transformation leaders can teach you about realistic artificial intelligence.”, California Management Review, Vol. 61, No. 4, pp. 110-134.

[26]「White Paper A Framework for Developing a National Artificial Intelligence Strategy」, https://www.weforum.org/whitepapers/a-framework-for-developing-a-national-artificial-intelligence-strategy/

[27] Stahl, B. C. (2022), “Responsible innovation ecosystems: Ethical implications of the application of the ecosystem concept to artificial intelligence.”, International Journal of Information Management, Vol. 62: 102441.

[28] Jabbarpour, M. R., Saghiri, A. M., & Sookhak, M. (2021), “A framework for component selection considering dark sides of artificial intelligence: a case study on autonomous vehicle.”, Electronics, Vol. 10, No. 4: 384.

[29] Littman, M. L., et al. (2022), “Gathering strength, gathering storms: The one hundred year study on artificial intelligence (AI100) 2021 study panel report.”, arXiv preprint arXiv:2210.15767.

[30] 「White Paper A Framework for Developing a National Artificial Intelligence Strategy」, https://www.weforum.org/whitepapers/a-framework-for-developing-a-national-artificial-intelligence-strategy/

[31] Valle-Cruz, D., et al. (2019), “A review of artificial intelligence in government and its potential from a public policy perspective.” Proceedings of the 20th Annual International Conference on Digital Government Research, pp. 91-99.

[32] Littman, M. L., et al. (2022), “Gathering strength, gathering storms: The one hundred year study on artificial intelligence (AI100) 2021 study panel report.”, arXiv preprint arXiv:2210.15767.

[33] Matjaž Perc, Mahmut Ozer, and Janja Hojnik. 2019. Social and juristic challenges of artificial intelligence. Palgrave Commun. 5, 1 (2019), 1–7. DOI:https://doi.org/10.1057/s41599-019-0278-x