Before predicting where AI could venture, it is crucial to understand the ways AI is shaping the world today.
According to the Singapore Computer Society:
“AI could be differentiated into Artificial Intelligence, Machine Learning, and Deep Learning.”
Artificial Intelligence
AI means to perform a variety of intelligence-related tasks, including planning, learning, manipulation, and problem solving.
The AI utilised today is considered weak AI, as we are only capable of executing single tasks within their designated usage and capacity. While strong AI encompasses the capabilities of a human being. This stage of AI is regarded as merely fictional. At the moment. We can see weak AI implemented in our everyday digital use, such as language translation apps, email spam filters, and facial recognition systems.
AI means to perform a variety of intelligence-related tasks, including planning, learning, manipulation, and problem solving.
The AI utilised today is considered weak AI, as we are only capable of executing single tasks within their designated usage and capacity. While strong AI encompasses the capabilities of a human being. This stage of AI is regarded as merely fictional. At the moment. We can see weak AI implemented in our everyday digital use, such as language translation apps, email spam filters, and facial recognition systems.
Machine learning is a part of artificial intelligence, where programs are trained to identify patterns in datasets and make decisions and forecasts based on those patterns.
Machine learning algorithms are used in GPS navigation systems such as showing congested routes based on the gathered data. Predictions like stock market prices and social media profile recommendations are also examples of application. Human intervention is still necessary when the results differ from the wanted solutions.
Deep Learning is a subset of machine learning with analytical abilities using usually 3 or more layers of neural networks.
Deep learning systems are able to figure things out on their own, such as autonomous vehicles and home assistant devices. They do not require human intervention, as the system is able to learn from its own mistakes. As a result, it requires large amounts of data and more training time, usually taking up to weeks. However, as technology continues to develop, the time needed to train the network could be reduced to hours in the future.
Impact on the workforce
As AI’s value increasingly lies in the ability to process complex information, more and more tasks require deep learning systems to create automated tools that could assist humans in making strategic decisions and digital transformation. We can start by defining four different types of intelligence and learn how AI has impacted the workplace.
Four intelligences in service tasks
According to Ming-Hui Huang and Roland T. Rust in Artificial Intelligence in Service
Mechanical Intelligence
Mechanical Intelligence requires minimal degree of training and education, and relies mostly on observations to respond. Jobs on this intelligence level include call centre agents, retail salespersons, waiters, and taxi drivers. These tasks are standardised, repetitive, routine, and transactional.
Analytical Intelligence
Analytical Intelligence focuses on logical, analytical, and rule-based learning. IBM’s chess player Deep Blue is a prime example of adapting based on data. Analytical Intelligence emphasises rational decision-making, and technical skills on data and analysis, such as data scientists, accountants, financial analysts, auto service technicians, and engineers.
Intuitive Intelligence
Intuitive Intelligence requires artificial neural networks, or deep learning systems, to learn and adapt based on understanding. The nature of these tasks are more complex and chaotic, and require intuitive, experiential and contextual interaction to solve problems. AI could assist jobs like marketing managers, management consultants, lawyers, doctors, and sales managers by providing personalised and experience-based services.
Empathetic Intelligence
The last level of AI is most advanced, and requires learning and adapting empathetically based on experience. Empathetic Intelligence incorporates social, communication, and relationship building skills, such as politicians, negotiators and psychiatrists. AI would need to solve tasks requiring empathy, emotional labour, or emotional analytics. Chat bots that communicate with customers and AI psychiatrists are current accomplishments on this intelligence level.
Corporations could start thinking about spending the money saved through automation on training employees for new jobs that could not be automated.
Need for regulations
- Haenlein, M., & Kaplan, A., 2019. A Brief History of Artificial Intelligence: On the Past, Present, and Future of Artificial Intelligence. California Management Review, 61(4), 5-14.
- Huang, M.-H., & Rust, R. T., 2018. Artificial Intelligence in Service. Journal of Service Research, 21(2), 155-172.
- Jenna Burrell, “How the Machine ‘Thinks’: Understanding Opacity in Machine Learning Algorithms,” Big Data & Society, 3/1 (June 2016): 1-12.
- Rozenfield, Monica. 2016. The next step for artificial intelligence is machines that get smarter on their own. The Institute. http://theinstitute.ieee.org/technology-topics/artificial-intelligence/the-next-step-for-artificial-intelligence-is-machines-that-get-smarter-on-their-own
- S.C.S.,2020. SIMPLIFYING THE DIFFERENCE: MACHINE LEARNING VS DEEP LEARNING https://www.scs.org.sg/articles/machine-learning-vs-deep-learning
- Tredinnick, L., 2017. Artificial intelligence and professional roles. Business Information Review, 34(1), 37-41.