By Sheraz Sharif, Head of Data & Cyber Security at Aion Digital
Humans are the most intelligent beings on earth, the ability to have consciousness, use tools, think about the future, and contemplate the past, makes us superior. We have the ability to understand, learn, and act. Now imagine amplifying the human potential via Artificial Intelligence (AI). The ingenious use of AI to improve our lifestyle is made possible through computational powers and data.
Artificial Intelligence is probably the most versatile and complex innovation made by mankind. It is defined in several ways, in layman terms, AI is the process of building intelligent computers using data to perform human-like tasks with speed, precision, and effectiveness. Although we see AI almost everywhere, it is a field that still vastly remains unexplored. At this point we are seeing just the tip the iceberg as the evolution of AI continues to grow with extensive impact on our lives and society. As per Gartner, organizations using Cognitive Systems in AI Projects will achieve long-term success four times more than others.
The concept of AI can be traced back to Allen Newell & Herbert A. Simon articulating a model of human and computer problem-solving in 1956. Alan Turing explored the mathematical possibility of artificial intelligence and then by John McCarthy in the same decade.
The idea of inactive objects coming to life as intelligent beings has been documented since the time of the ancient Greeks who referred to myths about robots. It went on to Chinese and Egyptian engineers building automatons in the year 400 BCE. However, the field of AI was not founded until 1956 during a Dartmouth College academic conference, in New Hampshire, where the term “artificial intelligence” was coined from a research project led by a small group of scientists. This was the birth of Artificial Intelligence and is what led to the mainstream version as we know it now.
AI is the use of machines to imitate human intelligence and is distinctly different from mechanical automation, which is a machine following a set of pre-defined instructions to accomplish a simple and repetitive task.
The truth is, AI is now used as a blanket statement for endless digital possibilities that it can offer. This often leads to confusion and hesitation to adopt as AI is applied and used in different businesses to address various target results. A quick spotlight on AI stages, types, and subsets should provide a clearer picture.
ANI (Artificial Narrow Intelligence): ANI is also known as Weak AI. It refers to AI systems that can only perform one specific task autonomously using human-like capabilities. It can target a single subset of cognitive abilities and only advances within its limitations. Examples of this would be Siri, Alexa, Google translate, image recognition, recommendation systems, spam filtering etc.
AGI (Artificial General Intelligence): Also known as Strong AI has the ability to learn, understand, and function like a human. It can independently build competencies and form connections and generalizations itself. AGI is just as capable as humans, however even with all the advancements and exponential increase of computing power, data, knowledge, and capabilities, we have not yet achieved the ideal vision of a Strong AI.
ASI (Artificial Super Intelligence): Will surpass human intelligence and can perform any task, this means it is no limited to replicating human beings. According to Forbes, this will be a pinnacle of AI research, as ASI will become by far the most capable forms of intelligence on earth.
Reactive Machines: Such AI systems do not store memories or past experiences for future actions. Reactive Machines specialize in just one field of work with focus on current scenarios and react on pre-defined tasks. Examples of this would be the famous IBM Chess program that beat the world champion.
Limited Memory: Collects previous data and continues adding it to their memory. Limited Memory machines can make informed and improved decisions by studying the past data from its memory. For example, it can suggest a restaurant based on location data that has been gathered. Another example are self-driving cars which use sensors to identify civilians crossing the road.
Theory of Mind: A more advanced type of AI that can understand thoughts, beliefs, and emotions, as well as interact socially. This category of machines is speculated to play a major role in psychology, however it has not yet been fully developed but research continues.
Self-Aware: This type is the future generation of AI. It will be intelligent and sentient, a machine with its own consciousness. Self-Awareness AI does not exist in reality and it is a hypothetical concept.
Machine Learning: A subset of AI that provides intelligence to machines with the ability to automatically learn with experiences without being explicitly programmed. ML collects and understands historical data, identifies patterns, and makes decisions. Examples of this would be forecasting, fraud analysis in banking, product & solutions recommendations, stock price prediction etc. ML is further diverged in Supervised learning (Classifications, Regression), Reinforcement learning (Positive & Negative Reinforcement learning) and Unsupervised learning (Clustering & Association).
Deep Learning or Deep Neural Network: Is an advanced level of ML, a process of implementing Neural Networks on high dimensional data to gain insights and devise solutions. The algorithms in DL are inspired by the structure and function of the human brain. Deep learning algorithms can work with an enormous amount of structured and unstructured data. Examples of this are self-driving cars, speech recognition, image recognition, automatic machine translation, etc. Some common deep learning types are Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Generative Adversarial Network (GAN) & Deep Belief Network (DBN).
Natural Language Processing/Understanding or Speech Recognition: Enables a computer to understand, process, and translate spoken or written human language into machine readable format. With the help of Natural Language Processing (NLP), we can instruct an AI system in our natural human language to communicate with a system. Examples of this are Siri, Alexa, Ok Google, Cortana. In addition, NLP is also used to detect offensive, profanity or abusive language, such as Twitter. Twitter uses NLP to filter out terroristic language in their tweets. There are two types of speech recognition: Speaker dependent and speaker independent
Machine/Computer Vision (symbolic learning): Has recently been getting a lot of attention and increasingly utilized AI science that enables systems to identify objects in the real-world, images and videos in the same way that humans do. Computer vision algorithms usually rely on convolutional neural networks (CNN). By 2022, the computer vision and hardware market is expected to reach $48.6 billion. Examples of this are self-driving cars, facial recognition, augmented and mixed reality, smart x-ray and MRI scans.
Robotics (symbolic learning): Robots are the programmed machines which can perform a series of actions automatically. The AI science for robotics focuses on artificial agents acting in a real-world to perform tasks with intelligence. AI and machine learning are being applied on robots to manufacture intelligent robots which can also interact socially like humans. Example: Sophia robot.
AI Adoption & Journey
AI is changing how organizations generate and utilize insights, especially financial institutions and their access to proliferation of data. As Fintech penetration is primarily focusing on AI, the finance sector is witnessing a revolution in its fundamental areas by radically shifting from the conventional ways of banking into smarter means of doing so. This is helping the industry optimize processes and find new revenue streams from back, to middle, to front office.
Given the endless possibilities of Artificial Intelligence, the use of AI in business continues to grow at an exponential rate with greater global spending year-after-year.
- As per International Data Corporation (IDC), Global spending on artificial intelligence is forecasted to double over the next four years, growing from $50.1 billion in 2020 to more than $110 billion in 2024.
- 60% of occupations will be automated, implying substantial workplace transformations and changes for all workers by 2030 (McKinsey).
- 25% of all digital workers will be using some form of conversational virtual assistant by 2022 (Gartner), with financial and insurance sectors leading the way (IPSoft).
- The global artificial intelligence market is expected to grow at a compound annual growth rate of 42.2% from 2020 to reach USD 733.6 billion by 2027.
AI is essentially shifting the paradigm of how we operate and do business with our customers. As AI becomes mainstream, we must realize the core benefits of AI and its adoption as part of a life-long journey and not as a technology component that is installed at once. This journey starts with an acceptance to Artificial Intelligence in our business (as AI novice), then progresses through phases to ultimately achieve the goal of AI-everywhere (as AI-first experts), deploying it throughout the organization and infuse it within the DNA of the business model.