Artificial Intelligence: Use and Potential Risks

Artificial Intelligence Uses

Artificial Intelligence (AI) technologies are nowadays utilized in a great variety of industries and sectors. For example, AI devices, such as machine learning and natural language processing, have become increasingly integrated into healthcare research. According to Jiang et al. (2017), the former type of device is primarily implemented to categorize various patient characteristics and forecast health outcomes, and the latter is applied to infer information from unstructured data and enrich research. Jiang et al. (2017) note that AI helps to deepen medical knowledge about leading causes of death, such as cancer and cardiovascular diseases, improve their diagnosis and intervention practices.

Notably, similar AI devices are implemented for research and analysis purposes in legal, education, and other industries. They are a crucial part of online marketing research and help retailers and manufacturers develop an in-depth insight into consumer preferences, sentiments, and motivations and consequently personalize marketing campaigns and final products (DHL & IBM, 2018).

AI is also integrated across a wide range of consumer-relevant activities, as well as logistics, manufacturing, and physical security practices. For instance, self-learning chatbots often serve to handle low-level inquiries across distinct industries, and some of them can even conduct relatively complex dialogs with customers (DHL & IBM, 2018). The use of AI to support autonomous transportation and production practices and their improvement by enabling the analysis of environmental data is a growing trend along with the utilization of facial analytics for security and other purposes (DHL & IBM, 2018; CB Information Services, 2017).

Overall, based on the review of all these AI uses and trends, it is possible to say that the discussed technologies mainly aim to enhance performance and time efficiency, personalize various services, and forecast distinct outcomes and events.

Benefits

Increased Analysis and Prediction Accuracy

AI devices assist in scanning large volumes of structured and unstructured data. Considering that it normally takes significant time and effort to analyze thousands of pages by using traditional means, AI allows increasing time efficiency and reducing the chance for human error. According to Mannino et al. (2015), when making decisions, individuals are prone to certain cognitive and psychological biases regardless of their competence and intelligence level. They may often show unrealistic optimism and underestimate risks especially if there are many different influential factors pertaining to a problem (Mannino et al. 2015).

It is evidence that judgment biases can lead to multiple detrimental outcomes and even be deadly in such spheres as healthcare. The making of such biases is not in the nature of AI technologies, but instead, they afford a rapid collection of data and recognition of complex information patterns. In this way, AI can offer significant advantages by supporting the jobs of decision-makers in different spheres of performance.

Increased Labor Cost-Efficacy

Automation and intelligent algorithms can significantly benefit business owners by inducing substantial savings. As stated by Wisskirchen et al. (2017), the major advantage of automated computer systems is that their performance does not depend on distinct external factors and, therefore, they are available at any time of the day. It is observed that the average cost of robot-assisted production per hour in the German automobile industry is about 5-8 euro, while the cost of a human worker is about 40 euro (Wisskirchen et al., 2017). For this reason, it becomes much cheaper to produce goods by using automated technologies than to outsource production to countries with cheap labor.

General Improvements in Lifestyle and Service Provision

AI can potentially increase efficiency in the performance of mundane tasks as well. For example, the integration of AI into households can lead to greater control over energy and water use, as well as enhancement of physical security at homes (Forbes Technology Council, 2018). However, it is possible to say that personalization of various services, such as education and healthcare, can have the most significant, positive impact on individuals’ lives.

As Alam and Kendall (2018) state, many universities already implement AI to personalize the learning experiences of students and note that such a flexible approach creates a more healthy and effective education ecosystem that recognizes diversity, takes into account individual strengths and weaknesses, and meets personal interests. In this way, it is valid to say that AI can lead to increased customer satisfaction and improved quality of life.

Major Concerns

Potential Negative Effects on Employment

The concern regarding the possible impacts of AI and automation on employment is widely discussed in professional and scholarly literature. Considering that the cost of AI-supported labor is cheaper than the work of human employees, employers may prefer the former because it will help to generate more profits. According to Huang and Rust (2018), at the current stage of technological advancement, AI can replace employees performing mechanical and analytical tasks.

It is suggested that in the future, AI will obtain more qualities of human intelligence and will become “softer” and more intuitive (Huang & Rust, 2018; Shabbir & Anwer, 2015). As a result, these technologies will be able to replace human labor entirely.

Nevertheless, some researchers and theorists argue that AI will never achieve a state that would match or surpass human performance. Still, its impact on the rate of job supply cannot be underestimated. Approximately 50% of jobs in the US market have a potential for automation (Frontier Economics, 2018). If full automation is under question so far, a radical transformation of jobs due to AI integration seems a more realistic possibility.

Notably, jobs requiring low education level, not involving complex interactions, and primarily based on manual skills are at the highest risk of replacement and transformation (Frontier Economics, 2018). Considering that these tasks are often performed by individuals who live in less privileged social-economic conditions and who especially need a stable source of income, work automation and consequent reduction in human job supply pose a significant ethical issue.

Security Threats

The increasing use of AI induces threats to security in the digital, physical, and political domains. According to Brundage et al. (2018), the new technology can be utilized to automate cybercrime tasks and lead to their greater efficacy and scale. It is expected that the growth in the rate of attacks exploiting “human vulnerabilities (e.g. through the use of speech synthesis for impersonation), existing software vulnerabilities (e.g. through automated hacking), or the vulnerabilities of AI systems (e.g. through adversarial examples and data poisoning)” (Brundage et al., 2018, p. 6). These observations emphasize the importance of protecting the digital information of users in the context of the increasing pervasiveness of AI technologies in human lives.

As for physical security threats, they are mainly related to automation. For example, the use of automated drones and other devices in the military sphere to perform attacks becomes a common practice, and the same approach can be used by individuals with malignant intentions against civilians (Brundage et al., 2018). Brundage et al. (2018) also note that attackers may disrupt such emerging cyber-physical systems as automated vehicles causing them to crash.

Lastly, risks associated with surveillance, deception, and persuasion by using AI-supported analytics and personalization tools are of great concern as well. For instance, “heavy personalization in Internet use can lead to the creation of personalized filter bubbles,” which increase the risk of targeted political messaging, propaganda, and exposure to different types of biased information (Osoba & Welser IV, 2017, p. 6). Thus, technologies that are created to benefit people can be utilized against them. Wrongful use of AI can compromise one’s privacy and expose him/her to data manipulation. All these risks to individual and public security indicate a need for supportive policies focused on AI implementation transparency, safety, and responsibility.

Recommendations

It is possible to say that the threats of both massive job losses and AI misuse can be partially prevented through the promotion of “a culture of responsibility” that would be created through user education, ethical statements, standards, policies, and norms (Brundage et al., 2018, p. 7). Specific policies should target employment issues and automated task performance. For instance, it may be recommended to protect those individuals whose skills will be considered obsolete as a result of task automation.

It is unlikely that all people will be able to afford to develop the necessary technical skills and knowledge needed to perform higher-order professional activities. Therefore, employers should be obliged to provide training and career opportunities that would enable employees at risk of job loss to maintain and improve their current standard of life.

Additionally, to mitigate security risks, policymakers should collaborate with technology researchers. Moreover, Brundage et al. (2018) suggest that the latter must take into account the potential for dual use of AI when they develop and enhance new technologies.

Accountable research is pivotal for the promotion of beneficial uses and prevention of harmful uses of distinct AI devices. Along with this, it may be recommended to identify the legitimate and illegitimate AI utilization practices and scopes within a policy. More importantly, governments and powerful private actors, such as multinational corporations, should be held accountable for AI implementation because they have the greatest power to influence the public by using these technologies.

References

Brundage, M., Avin, S., Clark, J., Toner, H., Eckersley, P., Garfinkel, B.,… Amodei, D. (2018). The malicious use of artificial intelligence: Forecasting, prevention, and mitigation. Web.

Shabbir, J., & Anwer, T. (2015). Artificial intelligence and its role in the near future. Journal of Latex Class Files, 14(8), 1-11.

Huang, M.-H., & Rust, R. T. (2018). Artificial intelligence in service. Journal of Service Research, 21(2), 155-172.

Osoba, O. A., & Welser IV, W. (2017). The risks of artificial intelligence to security and the future of work. Web.

Frontier Economics. (2018). The impact of artificial intelligence on work: An evidence review prepared for the Royal Society and the British Academy. Web.

Wisskirchen, G., Biacabe, B. T., Bormann, U., Muntz, A., Niehaus, G., Soler, G. J., Brauchitsch, B. (2017). Artificial intelligence and robotics and their impact on the workplace. Web.

Mannino, A., Althaus, D., Erhardt, J., Gloor, L., Hutter, A., & Metzinger, T. (2015). Artificial intelligence: Opportunities and risks. Web.

Jiang, F., Jiang, Y., Zhi, H., Dong, Y., Li, H., Ma, S.,… Wang, Y. (2017). Artificial intelligence in healthcare: Past, present and future. Stroke and Vascular Neurology, 2, 230-243.

DHL, & IBM. (2018). Artificial intelligence in logistics: A collaborative report by DHL and IBM on implications and use cases for the logistics industry. Web.

CB Information Services. (2017). Emerging AI: 7 Industries including law, HR, travel and media where AI is making an impact. CB Insights. Web.

Forbes Technology Council. (2018). 14 ways AI will benefit or harm society. Forbes. Web.

Alam, N., & Kendall, G. (2018). Five ways artificial intelligence will shape the future of universities. The Conversation. Web.

Cite this paper

Select style

Reference

StudyCorgi. (2021, June 16). Artificial Intelligence: Use and Potential Risks. https://studycorgi.com/artificial-intelligence-use-and-potential-risks/

Work Cited

"Artificial Intelligence: Use and Potential Risks." StudyCorgi, 16 June 2021, studycorgi.com/artificial-intelligence-use-and-potential-risks/.

* Hyperlink the URL after pasting it to your document

References

StudyCorgi. (2021) 'Artificial Intelligence: Use and Potential Risks'. 16 June.

1. StudyCorgi. "Artificial Intelligence: Use and Potential Risks." June 16, 2021. https://studycorgi.com/artificial-intelligence-use-and-potential-risks/.


Bibliography


StudyCorgi. "Artificial Intelligence: Use and Potential Risks." June 16, 2021. https://studycorgi.com/artificial-intelligence-use-and-potential-risks/.

References

StudyCorgi. 2021. "Artificial Intelligence: Use and Potential Risks." June 16, 2021. https://studycorgi.com/artificial-intelligence-use-and-potential-risks/.

This paper, “Artificial Intelligence: Use and Potential Risks”, was written and voluntary submitted to our free essay database by a straight-A student. Please ensure you properly reference the paper if you're using it to write your assignment.

Before publication, the StudyCorgi editorial team proofread and checked the paper to make sure it meets the highest standards in terms of grammar, punctuation, style, fact accuracy, copyright issues, and inclusive language. Last updated: .

If you are the author of this paper and no longer wish to have it published on StudyCorgi, request the removal. Please use the “Donate your paper” form to submit an essay.