What is AI? How is it defined, how is it applied and how is it being used across various industries?
August 4, 2023
Author: Samantha Booth
It seems that everywhere we look now there is a headline or article about AI. People are talking about it and are worried about AI and its implications for their future life and career. Most people can call to mind an example of a bad AI outcome. In Australia we are all aware of the Robodebt system, the consequences for the people involved and the continuing political fall out.
So what is AI?
Simply put AI combines computer science with huge datasets to enable problem solving. It incorporates machine learning and deep learning. AI algorithms are used to create systems which make either predictions, decisions or classifications based on the data that is inputted.
There are numerous definitions of AI. Here are just a few;
“artificial intelligence system” (AI system) means software that is developed with one or more of the techniques and approaches listed below and can, for a given set of human-defined objectives, generate outputs such as content, predictions, recommendations, or decisions influencing the environments they interact with: (a) Machine learning approaches, including supervised, unsupervised and reinforcement learning, using a wide variety of methods including deep learning; (b) Logic- and knowledge-based approaches, including knowledge representation, inductive (logic) programming, knowledge bases, inference and deductive engines, (symbolic) reasoning and expert systems; (c) Statistical approaches, Bayesian estimation, search and optimization methods. - EU AI ACT Definition
“It is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable." - John McCarthy from Stanford University in his paper “What is Artificial Intelligence”
“AI, or Artificial Intelligence, is a branch of computer science that aims to imbue software or machines with the ability to perform tasks that would normally require human intelligence. These tasks include things like problem solving, understanding natural language, recognizing patterns, perception, and learning from experience.” – Sourced from ChatGPT 4 on 4th August 2023
Key terms to understand when it comes to AI.
When you are reading or hearing about AI there will be lots of terms that you may never have heard before and probably do not have a good understanding of. The main terms that you will come across are listed below.
Machine Learning (ML): A subset of AI that involves algorithms and statistical models that enable machines to learn from data without being explicitly programmed. ML is a core technology that powers many AI applications.
Deep Learning: A specific machine learning technique that employs neural networks with multiple layers to process and learn from complex data, revolutionizing domains like image and speech recognition.
Neural Networks: A computational model inspired by the human brain's neural connections. Neural networks are the building blocks of deep learning and enable AI models to make complex decisions.
Natural Language Processing (NLP): The ability of machines to understand, interpret, and respond to human language. NLP is crucial for applications like chatbots and language translation.
Computer Vision: A field of AI that focuses on enabling computers to interpret visual information from images or videos. Computer vision is essential for tasks like object recognition and autonomous vehicles.
Robotics: The integration of AI and mechanical systems to create autonomous machines that can interact with the physical world.
Reinforcement Learning: A type of machine learning where an AI agent learns by interacting with an environment and receiving feedback in the form of rewards or penalties for its actions.
Supervised Learning: A machine learning approach where the model is trained on labelled data, meaning the input and the desired output are provided during training.
Unsupervised Learning: A machine learning approach where the model is trained on unlabelled data, and it tries to find patterns and relationships in the data on its own.
Artificial General Intelligence (AGI): A theoretical concept of AI that represents machines with human-like intelligence and abilities to perform any intellectual task that a human can.
Bias in AI: The presence of unfair or prejudiced decisions made by AI models due to biases present in the data or the algorithms used.
Algorithm: A step-by-step set of rules or instructions that a computer follows to perform a specific task or solve a problem.
Artificial Immune System (AIS): An Artificial Immune System (AIS) is a computer-based system that imitates how our human immune system works. AIS is designed to protect computer systems and solve complex problems. AIS is a useful tool for detecting and handling unusual or harmful things in computer systems and data, making sure everything stays safe and secure. It can recognize normal activities and patterns. AIS can also help in finding unusual things in large amounts of data. If there's something strange or unexpected in the data, AIS can detect it and alert us. Like our immune system learns from past illnesses and becomes better at fighting them, AIS can learn from past experiences and improve its performance over time.
Large Language Models (LLM): A Large Language Model is a type of AI model designed to process and understand human language on a massive scale. These models use deep learning techniques and a vast amount of data to generate text, answer questions, complete sentences, translate languages, and perform various natural language processing (NLP) tasks. The term "Large Language Model" indicates that these AI models are exceptionally big and complex, containing millions or even billions of parameters. Parameters are like the knobs or settings that the AI model adjusts during its training process to make accurate predictions or generate coherent text.
How could AI affect me?
While you may think that AI touching your lives is a new thing that’s not the case. There are several forms of AI that you’ve been using for many years.
Digital Assistants: Personal AI-powered assistants like Siri, Alexa, Google Assistant, and Cortana are widely used to answer queries, set reminders, control smart devices, and provide personalized recommendations.
Search Engines: AI algorithms refine search engine results, delivering more accurate and relevant information to users based on their preferences and past behaviour.
Social Media Algorithms: AI-powered algorithms drive content curation on social media platforms, ensuring users receive tailored content in their feeds, and they also gather user data for targeted advertising.
Online Shopping: AI is leveraged to offer personalized product recommendations, optimize pricing based on demand and supply, deploy customer service chatbots, and estimate shipping times.
Transportation and Navigation: Self-driving cars and real-time traffic management systems use AI to enhance safety, efficiency, and reduce traffic congestion, while AI-powered autopilots assist in aircraft navigation.
Text Editing and Autocorrect: AI algorithms are integrated into text editing software, providing spell-checking, grammar correction, and contextual suggestions based on user behaviour.
Gaming: AI systems play a significant role in the gaming industry, from providing challenging opponents in computer games to powering intelligent NPCs (non-playable characters) in interactive game worlds.
Fraud Prevention: AI algorithms analyse transaction data to detect suspicious activities and prevent fraud in financial transactions, safeguarding both businesses and customers.
More recently you will be becoming aware of ChatGPT, perhaps you’re using it. If you’re not using it yourself you probably know someone who is.
ChatGPT: Is a great example of a Large Language Model. The GPT stands for "Generative Pre-trained Transformer. ChatGPT is developed by OpenAI and is a language model with 175 billion parameters. It has been widely recognised for its ability to generate human-like text and perform a wide range of language-related tasks.
Large Language Models like GPT-3 are pre-trained on vast amounts of text data from the internet, which allows them to learn the patterns and structure of human language. After pre-training, they can be fine-tuned on specific tasks to make them more accurate and useful for particular applications.
Increasingly you will find that AI may be making decisions about you that will influence your life, such as will you gain admission to a particular university or be put forward for an interview for a job. This is where a lot of the recent concerns around AI come from. These AI systems are considered high risk systems and should be assessed, continually monitored and improved. Any such system must have a high level of human oversight as well as clear and accessible channels through which you can challenge the outcome and have it reviewed.
What Industries are Using AI?
Finance and Banking
Fraud Detection: AI is used to identify and prevent fraudulent activities in banking transactions.
Risk Assessment: AI algorithms analyse data to evaluate credit risk and determine loan eligibility.
Customer Support: Chatbots powered by AI provide instant support to customers for common queries and issues.
Medical Imaging: AI assists in the analysis of medical images, such as X-rays and MRIs, to aid in diagnosis.
Disease Diagnosis: AI helps doctors in diagnosing diseases by analysing patient data and symptoms.
Drug Discovery: AI accelerates the drug development process by predicting drug interactions and potential treatments.
Retail and E-commerce
Personalisation: AI algorithms recommend products and services based on user preferences and behaviour.
Inventory Management: AI optimises inventory levels and predicts demand patterns to reduce costs.
Supply Chain Optimisation: AI improves logistics and supply chain efficiency through predictive analytics.
Autonomous Vehicles: AI powers self-driving cars and assists in navigation and safety.
Traffic Management: AI optimises traffic flow and reduces congestion in urban areas.
Predictive Maintenance: AI analyses data from vehicles and infrastructure to predict maintenance needs and prevent breakdowns.
Marketing and Advertising
Targeted Advertising: AI analyses consumer data to deliver personalized ads and promotions.
Sentiment Analysis: AI gauges public sentiment towards brands and products on social media.
Content Generation: AI generates content for websites, social media posts, and marketing materials.
Personalised Learning: AI tailors educational content and resources based on individual student needs.
Automated Grading: AI can grade multiple-choice exams and assignments, saving time for educators.
Virtual Tutoring: AI-powered virtual tutors assist students in learning and problem-solving.
Quality Control: AI inspects products during the manufacturing process to detect defects and ensure quality.
Predictive Maintenance: AI monitors equipment performance and predicts maintenance needs to prevent downtime.
Process Optimisation: AI optimises production processes to increase efficiency and reduce waste.
Mining and Resources
Exploration: AI helps analyse geological data to identify potential mining sites.
Autonomous Machinery: AI is used in autonomous vehicles and equipment for mining operations.
Predictive Maintenance: AI monitors equipment health to prevent breakdowns and optimize maintenance.
Precision Farming: AI-powered drones and sensors collect data to optimize irrigation and fertilizer use.
Crop Monitoring: AI analyses satellite images to assess crop health and detect diseases.
Livestock Management: AI monitors animal health and behaviour to improve livestock farming practices.
These examples showcase how AI is used across various industries, enabling increased efficiency, better decision-making, and improved customer experiences. These are mostly benign and will have little impact on the average citizen other than some associated job losses.
We need to consider though that if we follow, as we tend to do, the USA and the UK we will soon have AI used throughout sectors such as policing, justice and border protection etc.
Predictive Policing: AI will be used to analyse historical crime data and patterns to predict and prevent criminal activities in specific areas. Qld Police have such a system being used in the domestic/family violence area.
Video Analytics: AI-powered video surveillance systems will help law enforcement agencies monitor public spaces, identify suspects, and investigate incidents.
Natural Language Processing: AI will assist in analysing social media and other online platforms for potential threats and criminal activities.
Facial Recognition: AI-powered facial recognition technology will aid in identifying individuals of interest in real-time. One such system was being used in the UK and was found to be very unreliable but the consequences for the people stopped on the streets, questioned etc was real – embarrassment, fear and anxiety.
Case Management: AI will be used to streamline and optimise case management processes, ensuring that cases are processed efficiently.
Legal Research: AI-powered tools will assist legal professionals in conducting research and analysing large volumes of legal documents.
Sentencing and Risk Assessment: AI algorithms may play a role in determining sentencing recommendations and assessing the risk of reoffending. These are already being used in the US and there is great concern that they contain algorithmic biases.
Biometric Identification: AI-powered biometric systems will be used for identity verification at border crossings.
Customs and Security Screening: AI will assist in screening and inspecting cargo, luggage, and passengers to identify potential threats.
Surveillance and Monitoring: AI-driven surveillance technologies will help monitor border areas for illegal crossings and suspicious activities.
It's important to note that while AI has the potential to improve efficiency and effectiveness in these areas, there are also ethical and privacy considerations to address. Ensuring transparency, fairness, and accountability in the use of AI technologies in any field but particularly in areas such as policing, justice, employment and education will be crucial to maintaining public trust and safeguarding individual rights.
If your company already operates AI systems or is thinking of purchasing or developing an AI system to aid in your business, it would be a good investment to have the system assessed from an AI Ethics and Governance perspective. Not only would an assessment ensure that the system is providing the best and fairest outcomes but it would help you to avoid breaching any of the existing discrimination, equal opportunity and privacy regulations/laws.