For the first time in history, mankind has discovered the innovation that can bring us closer to the futuristic world of human-technology integration. The development of Artificial Intelligence (AI) in the 21st century revolutionises the way we work and interact with machines and with one another. From flagging online content to creating autonomous vacuum robots, AI has had an undeniable impact on our lives. But what exactly is Artificial Intelligence? And are we truly ready for a new era of AI? Is AI a saviour or a saboteur to our digital society?
In this article, we will address these questions by journeying through the many changes of our digital society. Firstly, we will attempt to understand the meaning of Artificial Intelligence (AI). Afterwards, we will see how AI has expanded its role in different sectors of society from healthcare to the military. Consequently, we will observe how these new developments also brought new societal issues. AI development evokes new technical challenges in using the innovations in various fields. It also encourages ethical reflections on its impact on privacy norms, accountability, and the growing world inequalities. Lastly, we will argue how these challenges can be overcome by improving AI’s regulatory frameworks and creating a human-centric AI. To begin this journey, let us first take a look at our understanding of AI.
What is Artificial Intelligence?
Most of us are familiar with the term, but do we really know what AI means? The media term ‘AI’ usually refers to the computer science technique of Machine Learning (ML). ML originally started from simple experiments to test the ability of a computer checkers game (Machine Learning, 2021). Nowadays it is associated with any algorithm process that self-learns by obtaining greater data (Chattopadhyay & Majumdar, 2020).
Ever wonder how the online ads on your smartphone always seem to hit your jackpot of wishes? You have your software’s AI to thank (or to blame) for indulging your deepest shopping desires. Usually used for effective problem-solving, ML allows computers to analyse information and take decisions (Chattopadhyay & Majumdar, 2020). By creating an autonomous learning machine, production and research with limited time and resources more effective (Burns, 2021). For instance, one of its most common uses can be seen in recommendation engines. However, nowadays not only software engineers but also professionals in other fields make use of AI tools.
AI Occupying New Sectors of Society
Perhaps the most relevant and promising application of AI is in the field of public health. AI scientists are optimistic for AI’s help to improve human wellbeing and patient care by analysing healthcare data. AI can be used to alert healthcare professionals when deviations from normal health monitoring occur (OECD, 2019, p. 62). Although this application is far from widespread and experts still doubt its accuracy, the starting line has been crossed (DeGrave et al., 2021). In the future, scientists believe that AI can help detect and trace the potential spread of infectious diseases (Chattopadhyay & Majumdar, 2020). However, this does not mean that AI does not have current contributions to medicine.
In fact, nowadays medical professionals actively use AI to help with research and diagnosis. For instance, AI can help reduce the costs of drug discovery and help remote or visually impaired patients through image recognition (Chattopadhyay & Majumdar, 2020; Global Information Society Watch 2019, 2019, pp. 47-55). Referred to as precision medicine, AI processes patient’s records to alert professionals in making medical decisions (OECD, 2019, pp. 47-80). Thus, the development of AI in healthcare can bring a major lead for human well-being. Hopefully, our future grandchildren will never have to hear the words lockdown and quarantine ever again.
Aside from healthcare, do you know that AI also plays a role in our everyday food consumption? At first glance, we might not easily associate agriculture with high-tech AI systems. It is, however, one of the fields most benefitting from it. Moreover, its application is not all about imagined robot-farmers. For instance, algorithm learning allows farmers and researchers to obtain relevant weather and agronomic data (Chattopadhyay & Majumdar, 2020). Known as precision agriculture, it supports optimised crop production by gathering satellite data and using remote sensing devices (Precision Agriculture, n.d.).
Moreover, not only does AI lessens the workload, but it also reduces the inefficiency of human labour. The perfect example of this is present in a large greenhouse on the outskirts of Hanoi, Vietnam. Although indicated as a ‘vegetable factory’, there is no one working on its lettuce and tomatoes on a regular workday. The magic lies in the programmed fertilisers and water machines operated by Fujitsu, a Tokyo company. AI programming allows Japanese scientists to take care of the crops from their desks 4000 kilometres away from Hanoi (Global Information Society Watch 2019, 2019, pp. 20-24). As a result, this transformation reduces costs and brings greater profit and efficiency. It is therefore not overly ambitious to imagine AI as the next solution for the world’s food shortage.
Security and Public Safety
Another field that AI has strongly contributed to is public security. Law enforcement agencies around the world have increasingly used AI to tackle crimes. For example, to scope out areas and suspects for potential crimes. In other words, AI allows more efficient analysis of criminal data, producing a smaller monitoring target (Global Information Society Watch 2019, 2019, pp. 24-28). Additionally, another innovation that has been growing in its usage is the facial recognition surveillance system. Strongly embraced by governments, this software identifies a person by comparing video records with ID databases (e.g., mug shots, licenses, etc). Even when video footages are not available, sketches are enough to produce matching results. Therefore, the development of AI has been a major contributor to enforcing public security.
Additionally, AI also plays an ever-growing role in education. As the educational field explores the incredible possibilities of remote studying, AI is the perfect tool to help students develop. For instance, teachers increasingly use AI to assist with grading and learning. AI also allows the creation of personalised learning assistants that respond to students’ specialised needs. Most crucially, AI makes curricula accessible in diverse languages, improving learning for disabilities, internationals, and many others (Global Information Society Watch 2019, 2019, pp. 47-55).
New challenges to a more digitalised world
Nevertheless, as the saying goes, with greater power comes greater responsibilities. A more efficient, automatised world does not come without practical and ethical challenges. This is because of the common misconception that AI as an algorithm is neutral and without errors. However, it is important to recognise that the data used for the system still comes from a world far from equal (Global Information Society Watch 2019, 2019, pp. 9-14). Therefore, it becomes impossible to shield the algorithm from problematic data.
Consequently, this evokes a difficult technical task to expand the application of AI in diverse fields. For instance, AI operation in agriculture requires high-quality data to identify crops and leaves. This data collection is costly since scientists can only conduct research during the annual growing season. Another technical difficulty is the unequally distributed and often insufficient infrastructure to support the use of AI. Operational Machine Learning requires network connections and data warehousing systems, things that remain luxuries in many rural areas (OECD, 2019, pp. 47-80). For instance, basic healthcare and educational necessities are already taking a huge chunk out of public spending. Less developed regions will have trouble adapting to AI technology when they still struggle to provide the primary needs.
Moreover, aside from the technical challenges, the AI revolution also raises even more important questions on ethics. Firstly, researchers have expressed concern over privacy infringement. The development of AI can open up easier access to cyber-attacks and violations of ethical privacy. For instance, there have been critics over China’s use of the ‘social credit system’ in monitoring its citizens. Through surveillance, each Chinese person receives a score measuring their everyday behaviour from shopping to online habits. Consequently, this system encourages obedience as their scores heavily influence access to services and even the job market (Global Information Society Watch 2019, 2019, pp. 9-14).
Secondly, another ethical concern is on the subject of AI’s accountability. Machine logarithm blurs out the line of responsibility and hinders any appropriate rights protection mechanism. This consequently becomes dangerous in more serious situations with lives on the line. For example, modern warfare has shifted from sending ground troops to AI-controlled military drones. These Unmanned Combat Air Vehicles (UCAV) has received backlash in removing the humane accountability aspect in warfare. Its long-distance ability allows the military to cause collateral deaths without even pulling a trigger (Buchanan & Keohane, 2015). However, on a larger scale, there is an even more pressing problem.
AI: Increasing Inequalities?
Another significant consequence of AI use is that it widens the gap between the developing and the developed, the upper and the lower classes. Research has demonstrated that AI use supports the United Nations’ Sustainable Development Goals (SDGs). This is because technological and economic progress often go side by side. Concurrently, there is a constant race amongst states and private companies to chase the AI bus (Global Information Society Watch 2019, 2019, pp. 9-14).
This therefore means that a power shift is unavoidable. AI redistributes the dynamics between states and societies. The reason for this is because AI’s algorithm is deployed to and conducted from imperfect societies not immune to politics. Take, for instance, the crime pattern algorithm used by law enforcement. Although they can detect crime faster, it is common to find racial bias and discrimination in the process (Global Information Society Watch 2019, 2019, pp. 24-28). Therefore, on a societal level, AI can benefit those already privileged and worsen those less so.
On the other hand, on the international stage, the tension is between states with diverging levels of development. Large companies in the Global North are the original precursors of AI. Unfortunately, this means that other developing economies still struggle to gain access and adapt to the new technologies (Global Information Society Watch 2019, 2019, pp. 9-14). The reason for this is their different composition of the workforce. While human job replacement in the North is a blessing for its ageing working population, this situation causes extra problems for developing economies.
The majority of middle to lower-income countries are still in their ‘youth bulge’. For instance, the common age in Africa is 19,4 years and India is booming with the working population (Global Information Society Watch 2019, 2019, pp. 33-37). Consequently, issues of unemployment, increasing education and women labour rights are still the main focus for these countries. This of course does not leave much desire to invest in high-tech AI developments.
Moreover, although AI can be cost-efficient in the long term, it also invites troubles that these countries might not be ready to cope with. For instance, new technologies will centralise sectors such as agriculture to big industrial companies. This is a problem for emerging markets whose agricultural sector still rely on hands-on farming such as Southeast Asia. The replacement of human labour with machines will lead to major land and job losses (Global Information Society Watch 2019, 2019, pp. 20-24).
Therefore, despite the progress, the technical and ethical challenges of AI still hinder its widespread adoption. Although AI revolution promises efficient production and greater possibilities, not all states are ready for it. As a result, its major use is still clustered around several developed hubs, namely in North America and East Asia. What can we do to help these transformations? How can we make use of AI’s benefits without sacrificing other well-being factors?
The answer is to create AI algorithms designed to work with instead of for humans: Enhancing instead of replacing our capabilities. Firstly, this human-centric AI requires improved supervision and research. Governments need to be ready with better regulatory frameworks to prevent AI from damaging labour, land, or privacy rights (Global Information Society Watch 2019, 2019, pp. 9-14). Concurrently, this improved supervision will also encourage accountability and safe use.
Secondly, the AI machines themselves need to be changed. Although this might be difficult to imagine, the first seed of human-centric AI is already sowing. Countries such as India with strong informal markets have adapted their AI use by keeping humans in the loop. For instance, new AI platforms help organise blue-collared work in India and Indonesia. Apps such as Uber, Grab, Gojek delegate work automatically to bike-taxi drivers and delivery services. Furthermore, for other labour and logistic jobs, these apps allow the monitoring of work and rest time. It even serves as a mass recruitment agent through artificial conversation bots (Global Information Society Watch 2019, 2019, pp. 33-37).
Consequently, these algorithms make blue-collared work more efficient and rewarding, instead of serving as a replacement for their labour. Moreover, it also creates better working opportunities and well-being for the lower classes. A prominent example can be seen in India’s informal sector (cleaning, driving, domestic work, etc.). The society commonly associate these fields with workers from a certain caste, religion, and gender (Global Information Society Watch 2019, 2019, pp. 33-37). However, the platforms managed to break through this socio-cultural stigma online. Now people from all castes and backgrounds can find temporary work in the previously inaccessible jobs.
AI: From a novelty to an urgent need
In conclusion, we have witnessed how AI is helping to create effective progress in diverse fields from healthcare, agriculture, public security, to education. Nevertheless, implementing the AI revolution also comes with technical and ethical challenges. These tasks go beyond accommodating the high cost, data privacy concerns and accountability issues. More importantly, the crucial problem is the growing inequality gap to AI access between the developed and developing countries.
In this article, we propose a new approach: Creating human-centric AI that enhances our work instead of replacing them. By increasing research and creating better regulatory frameworks, we can monitor AI use to accommodate our work and prevent violations of rights. More importantly, human-centric AI will allow us to reach the middle ground. We will be able to pursue technological innovations while taking care of the world’s informal sectors and blue-collared workers.
As the world continues changing, this conversation between AI and human-centric issues becomes crucial. The day is approaching when AI is no longer a luxury but an urgent necessity, especially for those in the Global South. For instance, one of the most urgent problems in the world is the increasing pressure on the healthcare workforce. As the population continues to grow rapidly, societies are having a hard time producing healthcare professionals at the same speed. A prominent example, India produces only around 50.000 doctors each year against the country’s 1.36 billion mass population. As a result, each year its 1 million cancer patients have to fight for treatment from only 2000 oncologists (Sen et al., 2019).
A change is urgently needed. Computerised screening algorithms to perform mass diagnoses has become crucial, especially with lives on the lien. Therefore, AI can serve as an efficient intermediary between medical professionals and patients, providing quality healthcare at large.
Artificial Intelligence is one of humanity’s greatest innovations. Concludingly, this article outlines that the revolution towards a more automatised society is inevitable, yet crucially necessary. However, it is up to us to determine the path that AI will take our world into. Will we let it be a luxury for reckless uses of the privileged few? Or will we allow it to create a better, more equal world at large? The time to act is now, while the decision is still ours to make.
By Florencia Edgina Socio-political academic researcher and content writer.
*Write us if you are interested in the sources.