Artificial Intelligence: Benefits, Limitations, and UAE Government Initiatives

I. Introduction:

AI has been the subject of countless news articles, trending as a key topic on social media sites as well as government concern relating to its potential misuse.Below, we examine both the positive outcomes and limitations of using this technology and suggest ways for individuals and organizations to reap the benefits of AI while also mitigating risks. We further address the UAE government initiatives regarding AI and a possible future for AI.

II. Definitions:

What is Artificial Intelligence? AI refers to computer-generated systems capable of performing complex tasks that historically only a human could do, such as reasoning, making decisions, or solving problems. A simple definition is that AI consists of applications (apps) that perform complex tasks that once required human input. Examples of everyday AI at work include digital assistants (Apple’s Siri, Amazon’s Alexa); online shopping features (such as virtual assistants, dynamic pricing, fraud detection, and logistics optimization); chatbots (customer service); social media (Instagram, Facebook, X); electronic banking (transferring and depositing money from any location); and facial recognition software (on phones and laptops).

What is Generative Artificial Intelligence (GenAI)? GenAI is the latest version of AI technology that can produce several types of content, including text, imagery, audio, and synthetic data. These latest technological advancements are known as Content and/or Image Generators. These apps enable users to quickly generate new content based on a variety of inputs. Unlike traditional AI systems, which operate based on pre-existing data and rules, GenAI generates new data and ideas from digital art to novel text compositions.

ChatGPT4 (used for basic content creation tasks, edits, and brainstorming) is an example of a GenAI text content generator. There are Gen AI apps for almost every kind of writing. Examples include apps that create content based on a brand’s writing style and create fiction and edit novels. For legal practitioners, there are apps to review legal briefs and provide legal search assistance.

GenAI apps also create images from a text description or generate text captions from images. Such apps respond to a prompt which may be vocal or written by the user to create a photograph. GenAI can create written content as well as imagery based on the user’s request.

III. AI- Benefits and Limitations:

AI and GenAI are two powerful research and content generator tools that many businesses and people are embracing. Where AI performs labor-intensive work such as data gathering, organization of information, and content production, the resultant product frees up employees’ (those who previously did these tasks) time so that they may perform other more skilled work. In theory, improving business output and/or making social media communication more effective.

As with all technology, there are positive aspects as well as limitations. Foremost among the concerns is the inability of the user to identify the source of the resulting information. This makes it difficult to assess the bias of the underlying sources. This in turn affects the quality of results and content which may contain inaccurate or misleading information.

The resulting content can be biased when the app developer passes on unconscious biases during programming. For example, take an app used to compile and sort applications for employment. In its process, the app uses a person’s immutable characteristics such as gender, nationality, etc. to create its content. The resulting list of candidates could be based upon biases which could result in illegal hiring practices under the law—such as barring certain people from being interviewed and considered for employment. An organization using such apps would increase its risk of liability for claims of discrimination in its hiring process. It is critical that an organization that uses AI apps ensures have been tested and verified to be free from using illegal biases in its processing and production of content.

AI Image Generator apps generate images, photographs, and other visual images but have in some cases produced fake images (often termed “deepfakes”) and created picture/graphical representations called Hallucinations. A Hallucination is fake data that is statistically indistinguishable from factually correct data. It has sometimes created historically false photos and inaccurate images. AI Image Generator apps have caused problems for artists, photographers, and others trying to protect their content from unauthorized users. AI companies are aware of these issues and admit that GenAI is an evolving technology and is a work in progress.

AI has its limitations namely in producing content that may provide inaccurate or false information. Given the limitations, many AI users are correctly concerned about the lack of oversight of the development of AI applications and their potential for misuse. The ability to gather and process information at amazing speeds, or the ability to create content instantly with a few keystrokes, does not ensure the quality of the resulting data. As long as biased sources and people are putting out misinformation on the internet, there will always be a need for human oversight regarding AI results. As users of AI, it is imperative we be aware of its limitations and put in place a system of checks and balances to ensure we are responsibly using AI while reaping the benefits of its time-saving results.

IV. Current UAE Government AI Initiatives

The UAE’s goal is to become one of the leading nations in AI by 2031. Further, in 2017, the “UAE Strategy for Artificial Intelligence” was launched, which aims to:

• • Achieve objectives of UAE Centennial 2071.
  • Boost government performance at all levels.
  • Use integrated smart digital systems that can overcome challenges and provide quick efficient solutions.
  • Make UAE the first in the field of AI investments in multiple sectors.
  • Create a new vital market with high economic value.

This strategy will facilitate the creation of new educational, economic, and social opportunities for its citizens, government, and businesses through the implementation of AI technologies in various sectors (e.g. energy, tourism, and education). The AI Strategy will be implemented through the cooperation of multi-stakeholders, and various local and federal entities. Accordingly, the UAE benefits from having a robust foundation that is built on a diversified, multinational community, including global talent adept at exploring and testing AI solutions, and prone to adapting to new and advanced technologies at a faster pace.

The UAE government is additionally working towards accelerating the pace of AI adoption through strategic partnerships with both public and private sectors, and in this regard, has issued several documents addressing AI and its development, notably:

• • The National Program for AI Guide – showcases the UAE government’s commitment to becoming an AI technology hub.
  • The Smart Dubai AI Ethics Principles & Guidelines – an Emirate-level document issued under the Smart Dubai government initiative setting out a set of AI ethical principles and associated guidelines. A self-assessment tool was introduced together with the guidelines, to enable “AI developer organizations or AI operator organizations to evaluate the ethics level of an AI system” against the guidelines.
  • Generative AI Guide – operates as an explainer of AI and generative AI technologies and key challenges associated with them.
  • DIFC Data Protection Regulations – introduced detailed and specific requirements for those who develop, deploy, and operate AI systems.

These documents and related follow-up actions have been instrumental in establishing credibility in the UAE government’s aim to make the UAE an AI technology hub. For example, the Generative AI Guide ensures all stakeholders have access to information regarding Gen AI development and issues that arise. Similarly, the Smart Dubai Guidelines is an essential foundational document for developers and users to evaluate Gen AI content.

V. Future

There continues to be a range of perspectives on the potential of AI, presenting numerous likely future outcomes for AI in the UAE. There have been discussions focused on the job displacement that the increased digitalization of systems may bring about, including possible job losses and high levels of unemployment. On the other hand, this view has been challenged by several scholars (Mindell and Reynolds 2019, Thompson, 2020) with this perspective now evolving toward the quality of employment and the nature of work itself.  Moreover, as new forms of work models have developed, and more will follow with the increasing introduction of AI in the labor market, questions have arisen about the nature of employment.

For instance, the dynamics between self-employment compared to traditional employment, the terms and conditions of employment, and the relations between individuals performing the work and those who may own and oversee digital operations. It can be argued that digitalization has a dividing effect. It enhances efficiency by eliminating repetitive, routine tasks on one hand, but decreases low-skill job opportunities on the other. As a result, there will naturally be more of a growing need for specialized labor, preparing workers for emerging employment categories that prioritize specific, higher skill sets and adaptability.

A recent study for the World Government Summit in Dubai, in 2018, proved that for six Middle Eastern countries, 45% of the existing work activities in the labor market are automatable, given the use of current technologies, and that this risk is higher in sectors such as manufacturing and transportation, where routine tasks are performed. Additionally, 43% of existing work activities in the UAE are potentially at risk of being automated across various sectors – such as administration, manufacturing, government, and construction – and 300,000 jobs are also expected to be affected, including approximately 125,000 roles held by UAE nationals. Taking this into consideration, embracing digital competencies and basic AI knowledge will be instrumental in facilitating better decision-making within a labor market that is increasingly adopting automated technologies.

VI. Conclusion

The best practice for mitigating risks that using Gen AI may pose requires a multi-pronged approach that includes the government, Gen AI developers, and users. Initially, there must be a development of a culture that places major emphasis on ethical considerations with AI development and use. The core foundations of these ethical standards should be organized around transparency, accountability, and upholding societal values.

This approach requires the ongoing education of Gen AI developers, users, and government officials about this technology’s limitations and its possible consequences. As individuals, we must embrace digital competencies and basic AI knowledge. Ideally, the government would provide coordination and oversight of AI system development by providing ethical principles and associated guidelines for developers. Organizations that are using AI tools must monitor for potential misuse and check the accuracy of generated content using primary sources where available. Human oversight will always be necessary to manage and guide Gen AI development.

Fortunately, the objectives and subsequent UAE government actions in furtherance of the UAE Strategy for AI is a comprehensive approach. It addresses the limitations discussed in this article as it has initiated a long-term plan to put in place policies and initiatives that will guide future Gen AI development and meet the challenges to come.