Industry Updates

For a long time, artificial intelligence (AI) was restricted to university research projects and the R&D labs of technology vendors. However, in recent years elements of AI have begun to be integrated within smart devices and web services. Indeed, AI is becoming pervasive across businesses and is being used to enhance both internal and external services. According to IDC, global spending on AI is expected to reach $52.24 billion in 2021 .

The Middle East, Turkey, and Africa (META) region is often at the forefront of innovation when it comes to technologies such as mobile devices, the Internet of Things (IoT), and blockchain. Similarly, AI is being embraced by governments and organizations across the region as they look to create new services and improve their levels of efficiency. For example, AI sits at the core of the transformation ambitions outlined in Saudi Arabia’s Vision 2030 initiative, while the United Arab Emirates has established the UAE Strategy for Artificial Intelligence to tie in with the government’s ambition of enabling a superior quality of life.

AI is bringing about a new wave of transformation across industries, fueling demand for new types of skills sets, and driving dialogue around governance and ethics, and organizations must determine where AI can be used within their processes and identify the outcomes they wish to achieve.

Situation Overview

Over the past year, numerous debates and discussions have taken place around the use of AI, both globally and regionally. Typically, these discussions have revolved around the new types of research involving AI, the successful use of AI for business outcomes, the potential job losses that could be caused by AI, and the ever-popular topic of machines taking over humans. The major driver of these discussions is the fact that AI is no longer a concept limited to the research lab, with the technology becoming increasingly pervasive across consumer and business services.

What is AI?

In simple terms, artificial intelligence can be defined as activities devoted to making machines intelligent. Cognitive/artificial intelligence can be defined as “systems that learn, reason, and self-correct. The system hypothesizes and formulates possible answers based on available evidence, can be trained through the ingestion of vast amounts of content, and automatically adapts and learns from its mistakes and failures.” AI is already being utilized by many technology, ecommerce, and social media companies to either create a new service or enhance their existing services. Take, for example, the smart virtual assistants such as Apple’s Siri, Amazon’s Alexa, Microsoft’s Cortona, and OK Google that have been embedded within a variety of devices and systems. Other examples of AI already in use include facial recognition upon uploading an image to social media, recommendations of products on ecommerce sites, spam filters on email systems, and even being able to map optimum traffic routes during peak rush hours.

Several factors are driving the use of AI by organizations across multiple industries, including:

• Exponential Data Growth: According to IDC forecasts, global data volumes will reach 163 zettabytes (ZB) by 2025, up from 16 zettabytes in 2016. And there is a growing need to comprehend and analyze this data for strategic outcomes or real-time decision making. The use of AI will enable companies to analyze and manage their data much faster and across multiple iterations with minimum human intervention. This data can also be utilized to train AI systems for improved outcomes/services or for organizations to engage in deep learning.
• The Desire to Improve Productivity: Automating tasks to free up the time of knowledge workers so they can focus on more strategic and productive tasks is a major driver of AI adoption across many industries. This level of automation also helps address the recurring skills gap within organizations; knowledge workers can be retrained to undertake other tasks within their organization.
• Advancements in Technology: One of the biggest enablers for the use of AI is the accessibility to increased compute power at lower prices. This can be in terms of access to GPUs, cost-effective servers, and cloud services.
• All these factors have contributed to more and more organizations utilizing or exploring the use of AI within their organizations. AI is increasingly being considered as a major technology for the realization of digital transformation, which is when organizations use innovative technologies whereby to facilitate new operating models, create or enhance services, and gain a competitive edge to stay relevant in today’s hypercompetitive world. IDC’s META CIO Summit Survey 2017 showed that nearly 91% of organizations in the region are engaging in or planning to engage in digital transformation.
• Organizations across the region are engaging in pilots and proof of concepts (PoCs) utilizing various types of AI technologies and evaluating different types of use cases. AI is a broad term that encompasses many aspects such as machine learning, deep learning, natural language processing, image recognition, and recommender systems. The best way to comprehend this is to consider AI as an overarching term that incorporates machine learning and natural language processing. Then as a further subset of machine learning there is deep learning.
• Machine learning is the process of creating a statistical model from various types of data that performs various functions without having to be programmed by a human. Machine learning models are “trained” by various types of data (often, lots of data). Machine learning usually involves three types of learning (i.e., supervised, unsupervised, and reinforcement learning). Examples of machine learning include demand forecasting, recommender systems (used to provide suggestions within ecommerce sites), and fraud detection
• Deep learning is essentially in-depth learning or layers of learning and is part of machine learning. Examples for the use of deep learning includes autonomous driving, image recognition, video surveillance, and diagnostics.
• Natural language processing (NLP) is the ability to extract people, places, and things (also known as entities) as well as actions and relationships (also known as intents) from sentences and passages of unstructured text. It is inclusive of natural language understanding and natural language generation. Natural language generation is the ability to construct textual/conversational narratives from structured or semi-structured data. Examples of NLP includes sentiment analysis, question answering, and machine translation.

Different AI uses cases may, at times, use different elements of each of these technologies based on the automation or outcome that needs to be achieved.

Use Cases for AI

When it comes to the utilization of innovative technologies such as AI, it is critical for organizations to define the “use case” they are seeking to implement in order to ensure the right type of outcome. This approach essentially considers the business value created from the utilization of the technology rather than the technology itself. Several AI use cases are currently being utilized across various industries, including:

• Automated Customer Service Agents: The aim is to provide customer service via a learning program that understands customer needs and problems. It aims to reduce the time and resources spent in addressing customer queries and resolving customer issues. This is a popular use case across several sectors such as banking, insurance, retail, government, healthcare, telecommunications, and media. Example include chatbots on ecommerce sites or AI agents such as “Eva” being used by Emirates NBD in the UAE.
• Regulatory Intelligence: AI allows companies to more efficiently address their immediate regulatory compliance in real time by delivering actionable insights, limiting their exposure, and addressing issues as they arise. This use case is prominent across regulated sectors such as banking and finance, energy, and utilities. AI is also being used for anti-money laundering and fraud detection purposes.
• Program Advisors and Recommender Systems: In this use case, AI/cognitive capabilities are utilized to assist with user interaction or processing by matching the user’s needs to the right type or product or service. This is a use case being used by banks, retailers, governments, insurance firms, and telecom operators. Examples include the recommendations that are given to customers based on their online purchases and the way in which banks and insurance firms suggest suitable products/services after asking their customers to answer a series of questions.
• Automated Threat Intelligence and Prevention Systems: AI is used to process intelligence reports, extract critical pieces of information, and connect the dots between different pieces of information such as threats to databases, systems, website, and so forth. Examples include the use of AI for network and threat monitoring.
• Defense, Terrorism, Investigation, and Government Intelligence Systems: AI systems are used to help federal/state/local security services to identify, monitor, and respond to threats against personnel, assets, and infrastructure. Examples include using AI to enhance surveillance systems and enable identification at borders, as well as the use of robots for improving security in public places.
• Diagnosis and Treatment: This involves diagnosing conditions and enabling the provision of personalized treatment at the individual patient level by extracting insights from the intersection of diverse data sets, including medical records, lab tests, clinical studies, and medical images.
• Automated Preventive Maintenance: This system uses machine log data from various sources, contributing to a model that in turn will enable predictions and alerts around potential maintenance needs.
• Sales Process Recommendation and Automation: AI/cognitive engines work with the customer relationship management (CRM) systems to understand customer context in real time and recommend actions to the sales agents that are most relevant to the specific interactions in order to help them qualify or close a sale.
• Adaptive Learning: This system modifies the presentation of material in response to student performance. It also adapts trends in real time based on every interaction a student makes both during and in between lessons. Alef, an AI platform in the UAE, provides an interactive system that enables enhanced self-learning for students.

Digital Assistants for Enterprise Knowledge Workers: Digital assistants help workers answer questions, predict future events, and provide recommendations internal to the workplace. These intelligent systems leverage machine learning on large data sets, enabling innovation, collaboration, and higher employee productivity, thereby maximizing the return on information assets.

These use cases help organizations to optimize their processes, enhance their customer/user experience, ensure savings, and even create new products and services. AI will become increasingly pervasive in society at large with AI capabilities being included in consumer devices, public transportation vehicles, healthcare systems, education, and citizen services.

Adoption of AI in the META region

AI is a major transformational technology for organizations in the META region, with annual spending in this area expected to reach $156 million by 2021, which represents a five-year compound annual growth rate (CAGR) of 40.7%. The adoption of AI varies across the region, with countries such as Saudi Arabia, the UAE, and South Africa either already implementing certain use cases or putting in place strategies and long-term development plans for the adoption of AI.

The UAE has a strategic roadmap in place to drive the inclusion of AI across different sectors, and the country further highlighted its commitment to innovation by appointing a Minister of State for Artificial Intelligence, a first for any country in the world. One of the early AI use cases in the UAE was that of “Rashid”, a bilingual AI-based system or advisor in Dubai that serves as a single point of contact to guide users by providing information relating to various citizen and government services. AI customer service agents have since become commonplace in the banking and utilities sectors, while there are also examples of AI being used to dispense medicine, assist with emergency and crisis management, and streamline airline operations. In Africa, AI is being utilized in the banking and insurance sectors, while uses cases have also emerged in diverse areas such as wildlife preservation and remote medicine delivery.

In Saudi Arabia, AI is playing a key role in realizing the transformation ambitions of the Kingdom’s Vision 2030 initiative and driving innovation at large. The ambitious $500 billion Neom project will be entirely powered by AI, while AI, IoT, and cloud services will drive many interconnected services that will be availed by residents and visitors to facilitate a seamless and superior experience. And in October 2017, Saudi Arabia reaffirmed its commitment to the advancement of AI by granting citizenship to a “human-like” robot named Sophia.

AI-powered robots will undoubtedly have a major role to play in the future of the Kingdom, from being present at Neom to being utilized within the day-to-day operations of several key sectors. The Kingdom has further highlighted its commitment to innovation by investing in Japanese firm SoftBank – creators of humanoid robots such as “Pepper” and “Nao” that are being utilized across industries for customer service and information purposes.

Advances in AI will be critical for ensuring the successful integration of robots into society. These robots are different from manufacturing robots since AI will be used to teach them skills around human interaction and engagement. AI will be used to ensure that they can communicate in different languages and undertake many routine tasks. Early use cases of AI-powered robots are largely around the provision of customer services, patient care, and general assistance.

It should be highlighted that these are essentially physical robots. These are different from “software” robots, also known as robotic process automation (RPA). RPA is software that tends to imitate human actions and is used for automating many mundane tasks. It is based on rules that are provided to the system. AI goes much further than this, encompassing rules-based systems and ensuring that the machine can learn and respond from humans. It is also much faster and able to respond to unpredictive scenarios. Increasingly, many RPA vendors are enhancing their functionalities with AI.

This demand for AI is creating a new ecosystem that encompasses AI/cognitive-based process and industry applications, AI/cognitive-based business to business processes/services, and AI/cognitive-based consumer services. This drive to adopt to AI/cognitive solutions has led to the establishment of several startups in the region that are fueling innovation and economic diversification in the wider economy.

Challenges Around the Adoption of AI

The challenges around AI largely relate to the way in which data is going to be managed and utilized to train the system.

• Bias: Organizations need to ensure that the data being used to train the system has as little bias as possible. Clean and structured data sets are a necessity to ensure the right type of response to the system. Any bias that is introduced to the systems can create unexpected and unwanted outcomes.
• Privacy: One of the biggest use cases of AI is for enhancing consumer services and engagement. This requires organizations to ensure data privacy and security when harvesting data to provide these services. It is critical for organizations to ensure they are adhering to global data privacy standards such as the European Union’s General Data Protection Regulation (GDPR). These regulations provide customers with greater control over their data and the way in which it can be used to improve overall transparency in the utilization of services.
• Trust: In a world full of autonomous services, trust becomes critical. Driverless cars bring this issue to the forefront. If the car does not respond on time and in the right manner, it can be fatal. Users need to be able to trust that the system will respond to incidents properly and instantly.
• Skills: Skills are a major challenge for customers, vendors, and partners. Technology vendors in the region are working with governments and education institutions to train students around AI. Partners and customers need to consider retraining existing employees around AI as well.

Telcos and AI

Telcos will play a major role in ensuring the realization of many AI use cases. Even IoT projects that are further enhanced with AI services require telcos to play a major role. These are projects such as driverless cars, intelligent transportation management systems, and the use of drones for preventive maintenance. In order for them to be implemented effectively, these projects rely heavily on telcos that can provide superior connectivity services such as 5G.

Increasingly, telcos are utilizing AI to enhance their own customer and network services, such as using chatbots to engage with customers and utilizing AI to enable self-healing and secure networks. Self-healing networks can maximize operational efficiency and choose the right course of action to ensure uptime without the need for human intervention. The utilization of AI within telcos also improves the overall customer experience.

Increasingly, telcos are diversifying their services portfolios and becoming major IT services partners to their enterprise customers. With AI, telcos will be able to not only support their customers in the implementation of AI services, but also provide platforms that enable their customers to use AI services.

Telcos stand to clearly benefit from AI, both in terms of optimizing their internal operations and providing enhanced services.

STC also is committed to provide such enhancement to better serve its customer base. STC is one of the largest infrastructure, telecommunications, and digital services providers in the Kingdom of Saudi Arabia. Voice and mobile data volumes from their subscriber base and data being generated from IoT deployments can be leveraged and enhanced with AI to provide STC’s customers with strategic insights and customized services. Projects that have been undertaken by STC include working with the Ministry of Tourism to help them better understand visitor behavior and trends in the kingdom. Analysis of visitor data can be utilized to provide visitors with customized services such as helping them plan their journey, offers of routes they can utilize, and activities they can pursue at their destination. STC is also working with the Ministry of Transportation in the area of traffic management, and AI can be used to further enhance these services. Other use cases that can be augmented through the use of AI include customer insights, sports analytics, epidemic control, and crime investigation.

Conclusion

AI will no doubt lead to a new wave of innovative and superior user/customer experiences. But despite the apparent benefits of AI, many people still equate the proliferation AI to job losses and, in extreme cases, to robots taking over the world. Just like other technologies, AI will cause a shift in the type of jobs that people perform rather than creating widespread lay-offs. Individuals with skill sets that are no longer required by organizations will need to be retrained; as such, AI will create demand for new jobs such as AI trainers, Machine interaction modelers and business analysts that can ensure the seamless integration of AI for business functions.

As AI continues to advance, popular media tends to highlight the issue of robots taking over from humans, but this is extremely unlikely to happen. The term “singularity” describes when AI is either equal or superior to humans but achieving this level of utilization will require huge advances in computing power, algorithms, and supporting regulations. While such advances may be possible in the very distant future, it is not something that is even a remote prospect for the near term.

The way organizations deliver services is clearly evolving. In many cases, customers do not realize they are engaging with AI systems, but the technology has come to define customer expectations. Across the META region, AI has already disrupted the way in which customer services are provided by banks, telcos, and government entities, while image recognition is being utilized for security purposes and recommender systems are being used to enhance ecommerce services. The benefits of using AI are clear to see, and if organizations are to achieve their digital transformation goals and remain relevant in a changing digital economy, they must embrace a proactive approach to identifying which AI use cases will be most beneficial to their particular circumstances.

Source: http://businessweekme.com/the-rise-of-ai-in-the-digital-era/