What is AI?
As one of the most prevalent and transformative technology of all time, Artificial Intelligence (AI) has been incorporated in various major industries since the term was first coined in the 1950s. From providing services such as virtual shopping and online banking to reducing overhead costs in manufacturing and streamlining diagnostics in healthcare, AI has driven most industries forward and has unquestionably changed lives of many.
Al is regarded as the continuous understanding and development of computer systems to perform tasks. The three categories of AI are: Artificial Narrow Intelligence (ANI), Artificial General Intelligence (AGI) and Artificial Super Intelligence (ASI). In ANI, AI is programmed to mimic human’s cognitive abilities, however its abilities are set within a narrow range of parameters. Examples of ANI include Google’s Rank Brain and Apple’s Siri. In these examples, ANI are able to identify, rank and filter out information in terms of relevancy based on keywords provided.
AGI is created to mimic human behaviour and intelligence to perform labour and cognitive tasks at a pace that is significantly quicker than human, for instance online chess games and machinery in the production line.
On an extreme side of the spectrum, ASI is programmed to surpass human’s abilities in terms of task performance and have capabilities that are beyond human’s imagination. In contrast to ANI and AGI, ASI is yet to be developed and remains in the research and experimental stages. In fact, AI that actually surpasses human’s cognitive abilities remains a myth to this day because no scientists are completely confident with the possibility of ASI. However, according to a Forbes article – ‘The Future of AI’, AI startups that aim to develop ASI have reached a record high in venture capital funding. In 2018, AI startups have collectively managed a $9.3 billion for AI research and development, an increase of 72% from 2017’s AI-related funding.
AI start-ups aside, well-established key players in the AI industry, to name a few, are Google, Baidu, IBM, Microsoft, SAP, Intel and many others. USA currently takes the greatest market share, seizing 44.52% of the global market, followed by Europe, taking up 19.43% market share.
The Mechanism of Artificial Intelligence
With increasing data volumes, improvements in computer storage and advancement in algorithms, AI technologies are driven by data input and human inquiry. Based on the data input, AI adapts to the information provided by identifying regularities and patterns, whilst automating the processes of self-learning. AI utilises a method known as back propagation where it continuously adjusts to the addition of data input until an accurate answer is achieved. As AI performs deep learning and automates advanced algorithms, the quantity and quality of input data is key. For instance, Google Photos, Search Engines and Siri achieves greater accuracy when they are more frequently used. In a competitive environment where all users input the same amount of data into their computer systems, the programme with the highest quality of data will outperform the rest.
Essentially, AI technology is embedded with numerous functions: machine learning, deep learning, cognitive computing and natural language processing. When data is fed into the computer system, data processing will be established. A neural network must form to process the data. The network consists of interconnected units that process input information, propagating from one unit to another. The stage of neural network formation will acquire an iterative algorithm to form connections within the pool of information. When such connections are made, machine learning is automated where the system is able to derive meaning of the data and implicit insights without receiving any direct and explicit questions. Deep learning is similar to machine learning, utilising neural networks to derive complex patterns in data. Machine learning provides analytical solutions, whilst deep learning provides visual and sound recognition.
Cognitive computing is the ability to mimic human capabilities, for instance, analysing visual and sound inputs, forming coherent sentences and ultimately providing human-like interactions. Natural language processing is the ability to understand, recognise and analyse human languages. With such ability, humans will be able to communicate with the system via language input into the system.
These functions are only enabled with the following supporting technologies: graphical processing units, the Internet of Things and advanced algorithms. Graphical processing units provides significant computing power to make neural networks and data storage possible. The expansion of the Internet of Things has made available endless amounts of information and data for input purposes. Advanced algorithms create the possibility of iterative methods, back propagation and information relay in computer system.
Artificial Intelligence – Government Use
AI technologies may potentially spearhead significant changes in the day-to-day management and operational affairs of the government. AI can aid public interactions with the government, for instance, the introduction of virtual assistants and chatbots in outlets and call centers. These allow public queries to be answered efficiently, as well as preventing long queues and waiting time. AI can also be introduced in a series of government services such as directing complicated queries or requests to appropriate department, assisting members of the public in filling out forms and scheduling appointments. In 2016, the Australian government incorporated virtual assistants in its taxation office, national disability scheme and department of human services. AI technologies can be used to allocate resources efficiently – providing administrative support and carrying out repetitive tasks where needed.
With the enormous amount of data stored in the government database, AI is capable of organising and combining numerous datasets to derive insights information, as well as summarising a variety form of data.
Artificial Intelligence – Facial Recognition (FR)
Computer vision, a study of AI, has made Facial Recognition (FR) mechanism possible by measuring up to 80 nodal points, distances between points on faces. The AI is then able to store the nodal points distances in its database and match the name of the individual to its data. To build the FR mechanism, a neural network is formed in the system by an input of data to generate pattern recognition, and these data consists of face images of millions compiled from social networks, websites, surveillance cameras and dating apps.
Due to its high social acceptability and minimal error of 0.8%, FR mechanism is widely incorporated in Snapchat’s animated lenses, Apple’s Face ID and security identification. FR is unlikely to be limited to these applications, as technology companies are working on FR to provide these services to banks, ICOS and e-businesses.
In 2017, FR took over the F&B industry by storm when KFC in China allowed customers to pay via FR technology and when frozen pizza brand DiGiorno uses FR to capture customers’ expressions when eating the pizza in an effort to plan its future marketing campaigns.
Artificial Intelligence – Crowd Counting
Using drone, aircrafts and satellite, images and videos of crowds are taken from afar and processed in the computer systems. The neural network is formed when a large amount of training data is imported into the system. This allows the system to be able to distinguish stagnant or moving human or objects based on colour, shape and geometry pixels. The method tracks individuals and prevents double counting, thus achieving an incredible accuracy.
During the Hong Kong protest marches in June 2019, approximately 600,000 headcounts were made using AI technology. Instead of using cameras on the air, seven iPads were placed at two stations to record the crowd movement and head counts were made when people cross a benchmark line in the frame of the video. Prior to the protest, a trial crowd counting was run to input data such as density of people, direction of crowd flow, speed of crowd flow and lighting conditions, these allow the network system to adapt to the conditions of the built environment.
Previously, infrared beam counters and thermal counters were used to count people by detecting sources of heat in human body. These technologies have achieved an average accuracy of up to 80%. With the current technology of video and Wi-Fi counting, 90% accuracy is achieved.
Artificial Intelligence – Self-Driving Cars
Another notable leading-edge technology is self-driving vehicles, the use of these vehicles is yet to be widespread as such technology is still in the stages of road testing. These autonomous vehicles are powered by AI deep learning, where sound and images data are fed into the network system. A series of functions such as voice and speech recognition, image recognition and processing and motion detection can be performed, thus allowing these vehicles to be able to recognise all possible traffic situations and perform the right action at the right time.
According to Allianz, the global automated car market will arrive at $43 billion by 2025. Companies that produce chips, sensors, software and laser scanning will be most in need during the next decade. Electric battery is likely to replace petrol, and thus reducing the emission of greenhouse gases. Self-driving cars are believed to be safe and smart, as it will incorporate many automated features. Sensors are featured to help detect obstacles by reflecting pulses of lights off the car’s surroundings. Video cameras will be installed to detect traffic lights, read road signs and keep track of other vehicles and pedestrians.
Artificial Intelligence in Malaysia and ASEAN – The Legal System
An example that draws closest to the hearts of Malaysians are the implementation of AI in Malaysia’s legal system. Earlier in March this year, Malaysia’s Chief Justice Tan Sri Richard Malajum (now retired) announced the introduction of AI to assist judges in making decisions on punishments for convicted criminals and reducing disparity in similar crime sentences. There was a strong belief that the AI will prevent unfairness and will stand by a set of values such as credibility, accountability and transparency. As such, AI was strongly believed to have complemented the legal service industry.
In addition to ensuring unbiased punishments, AI plays a significant role in the legal profession in Singapore. Legal firms in Singapore have leveraged on technology to assist them in document management, identifying and browsing through relevant case studies, legal analysis and research.
The Future
AI technology has yielded great success in various industries, however its peak of development has yet to come. In 2016, AI’s global market was worth $4 billion. In the 2019 Global Industry Analysis & Forecast, the report states that the AI market is predicted to reach $169 billion by 2025 and multiply by approximately 100 times to $15.7 trillion in 2035. International Data Corporation (IDC) estimates that retail, banking, healthcare and manufacturing will make up half of the AI-related fund.
Written by Jenn Tan, Intern at 27 Advisory, studying Civil Engineering at Imperial College London. Love exploring new places, taking up new experiences and living in the moment. Enjoy learning about current and future technological trends and their impact in society advancement, thus researching about artificial intelligence and its applications in various industries.
Having more than 27 years in business, 27 Group is able to provide you with the access to investors for competitive funding needs while providing better ways to operate your business through financial and corporate advisory. We are the only 100% Malaysian owned local consulting firm that is fast, flexible and focused with unique expertise that blends of local socio-economic policy setting, engineering built assets globally and detailed in financial analysis.
We do project development integration to improve project returns and are committed to providing a sustainable environment for a better tomorrow. Our delivery model blends values important to humanity into business strategy through socio-economic transformation modules and we are passionate about building opportunities for the next generation to achieve their highest potential.
#rebuildinghumanity is 27 Group’s vision to collectively rebuild our nation through assets we build (eg. infrastructure, real estate, hospitals) and natural capital (gas resources, plantations, human talent) using innovative and sustainable methodologies.
Speak and rebuild with us if you need fresh ideas or more efficient financing or project implementation to improve your KPIs: ivan@27advisory.com.my
