ai literature review

ai literature review

The Impact of Artificial Intelligence on Society

1. Introduction

At the same time, an AI with the capability of self-improvement could be an unpredictable threat. This risk has led to efforts to devise a process for the control and testing of an AI before it reaches the level of human intelligence. Some forms of narrow intelligence are more desirable than others. A machine which is hard to turn off seems intelligent because humans don’t turn off a conscious being without a good reason. A medical decision support system designed to help doctors diagnose diseases should make it easy to switch off the recommendation AI when it is doing something risky. There are grave concerns for a misaligned AI that does not do what we ask but follows the letter of the instructions. For this reason, it is essential that the design of a new intelligent system takes into account the motivations and goals of the beings it will interact with. Finally, it has been suggested that an international shared commons should be created for the sharing of beneficial AI to prevent destructive competition and reduce the risk of an unbalanced distribution of AI capabilities.

Artificial intelligence has the potential for aiding humanity in several ways. Simulations could be run in virtual reality to test a hypothesis and the effect of an experimental variable upon an autonomous agent can be observed. This is vastly more cost effective and efficient than the present method of trial and error using living organisms. In a similar vein, an AI can be observed in a safe environment to test potential harmful capabilities. Numerous systems can be taught from the cumulative experiences of many trainers (anomalies may be detected and corrected en route) and best practices can be disseminated throughout the AI population, resulting in a net increase to the sum of desired outcomes. Finally, the creation of an AI which is smarter than humans has the potential to solve some of the world’s most pressing problems.

2. Applications of Artificial Intelligence

Applications of AI have been used in a wide range of fields. Medical diagnosis, stock trading, robot control, law, scientific discovery, and toys are just some of the fields AI has been used. Having an enormous impact in various fields, AI has the potential to create more artifacts. On the other hand, AI also offers the potential to significantly and fundamentally alter aspects of society. There are various ways in which researchers in AI and AI applications can build in ethical and social considerations into their work. For example, as AI artifacts often have the potential to significantly affect the well-being of individuals and society, it is important to ensure that such artifacts are designed and tested to be reliable, secure, and robust. This will often require building in considerations of social and ethical issues at the very beginning of a project and at all stages in the development of the artifact. This contrasts with the typical approach where consideration of social and ethical issues only comes at the end of development, usually spurred on by the recognition of a possible negative consequence of the artifact. AI and AI research can also seek to build in considerations of social and ethical issues in the development of an artifact by bringing together multidisciplinary teams that include social scientists and ethicists. This can also involve the direct involvement of stakeholders and the end users of the artifact. As the saying goes, the best defense is a good offense, and this direct consideration of the broader implications of an AI artifact can be the most effective way of avoiding a harmful impact. These general issues with direct relevance to the design and development of AI artifacts rightly deserve specific attention in their own right. AI has the potential to be used as a tool to address some of these issues, for example by the use of AI in economic modeling to better understand the causes of global economic disparity or through the design of decision support systems to help policymakers make better decisions.

3. Ethical Considerations in AI Development

An area with high hazard and little understanding involved in decision making is that of public policy formulation. On the surface, public policy is aimed at the best solution for the present and future of society as a whole. AI offers the prospect of simulating the effects of a certain policy or exploring the possibility of alternatives under simulated environments. This can be achieved via agent-based models in an emergent and convoluted environment or through the solution of models constructed from data using machine learning techniques. While this could be a great enabler toward greater understanding of complex systems and better policy, there is great potential for catastrophic failure. It may be that a policy has one outcome in the short term and another in the long term. A simple example would be a climate policy aimed at reducing global temperatures in 100 years’ time. The second, third, and nth order effects of such a policy are highly unpredictable and necessitate a simulated environment to avoid doing more damage than good.

Today, most of us are sentient in the knowledge that our society is poised on the crest of an epoch-making change when we will be at the sharp end of a vastly complex world where decisions will be ever more significant and widespread in their impact. Societal impacts of AI are wide-ranging, encompassing complex employment and economic influences. They will sometimes be subtle and subliminal yet have pervasive steadfast effects in the long term. AI, peradventure the ultimate tool in rationalization banner toward explication of problems and a science to instruct the rationalization process, may have a momentous impact on the methodology of decision making in society.

4. Benefits and Challenges of AI Implementation

There are many types of automation, and not all are equally valuable. The most basic automation is a scripted sequence of actions. Often this can be done using intelligent searching and data mining to find the data to act on. More valuable is the automation where the program is given a goal, some data, and then calculates the best course of action to achieve that goal. The ultimate automation is to remove the human from the loop and have the program act on the data in the most general sense. An example of the last type of automation is a self-driving car. The car is given the goal of getting from point A to point B. It is presented with various data such as GPS coordinates, maps, and camera data and must make many decisions on the best way to achieve the goal based on that data. In removing the human from the loop for such tasks, AI has the further benefit of not only automating an existing task but also providing an alternative method for that task which can replace the human.

At the implementation of AI, the greatest benefit is to increase automation. Automation is an important target of the explanation of AI because it is how learning can be said to be applied to a new task. What is automation? It is the step from being instructed how to do something to learning how to do something. Often, in the computer program world, it means the program is data-driven. The program is presented with data in a particular format and then is tasked with performing operations on that data. In general, we can say that automation is a movement in the direction of AI when a program has the ability to analyze the data and make decisions based on that data.

5. Conclusion

The society is still to experience greater revolutions and impacts in the coming years due to AI. The rate at which technology is advancing, it’s evident that within a span of years, artificial intelligence will shape up the future of any society in all parts of the world. However, given AI’s long-term potential, there is no turning back from here. It is a critical time for humanity. Society will have to bring balance in the further advancement of AI within various sectors (such as economic, health, and social). Steps will have to be taken towards understanding AI’s long-term impacts to ensure that it does more good than damage to society. AI’s capability to be at parity with human intelligence would require redefining of the very basic tenets of modern society. The cause-effect for all consequences of AI modes would have to be thoroughly understood in the present context so seek that we are not caught off guard. This is a tool that will take the human society to a new level, and at this point in time, it is difficult to fathom if it will be for better or worse.