What is AI? Understanding Artificial Intelligence in Simple Terms!
Introduction to Artificial Intelligence (AI)? : Learn, Discover, and Take advantage of AI systems.
Introduction
Hi, have you imagined systems capable of autonomously driving cars, completing errands, performing surgical operations, managing logistics, and various aspects of human endeavors, technological innovation and how it can be achieved? In this article, I'll simplify artificial intelligence (AI), covering its applications, categories, types, technologies, benefits, risks, and offer additional resources for further learning.
At the heart of every endeavor in computing and mathematics lies the pursuit of creativity and innovation, driven by the intention to enhance human lives in various ways. This motivation comes from the fact that things always change, and people's needs keep expanding.
For example, embarking on a journey typically requires us to steer and navigate towards our destination, regardless of our physical condition. However, AI-powered cars can autonomously drive to the same destination. Unlike humans, AI systems do not experience fatigue or exhaustion while performing tasks.
To keep things straightforward, I'll skip history of AI. While various articles and publications may offer differing perspectives, notable figures commonly associated with the conception of AI include Alan Mathison Turing and John McCarthy.
What is Artificial Intelligence?
Computer systems built with the ability to carry out complex tasks that traditionally could only be initiated and completed by humans, such as learning, thinking, perceiving, making decisions, or solving problems, are referred to as artificial intelligence (AI).
Definition of some of the cognitive abilities of AI
General learning: the ability of machines to learn from data, recognize patterns, and make decisions independently, without needing specific instructions
Reasoning: the capacity of artificial systems to process data, decide, and resolve issues using rational rules and principles.
Problem solving: the ability to use sets of rules, guidelines, and thinking methods to handle difficult tasks or problems.
Perception: the ability to interpret and understand information from an environment using sensors, cameras, or other input devices.
Language: the ability to understand, create, and work with speech or written language
Categories of AI
In some research or publications there could be mention of two categories Weak and Strong AI while others mention three categories. Let's take a look at them:
Weak AI (Narrow AI): AI systems that are designed or constrained (limited) to perform specific tasks.
Strong AI (Artificial Generative Intelligence AGI): AI systems that perform tasks with similar human cognitive abilities such as learning, reasoning, perception, understanding and problem solving techniques. Under development.
Super Intelligent AI: This represents a more advanced form of AI that researchers are currently developing. It would entail understanding and remembering emotions, beliefs, needs, and depending on those, making decisions.
Types of AI
Below is an overview of each type of AI, as outlined by Professor Arend Hintze from the University of Michigan.
Purely Reactive: these are AI systems that do not keep track of previous experiences to perform operations on them in the future.
Limited Memory: these AI systems can collect previous experiences and keep track of new experiences progressively.
Theory of Mind: these AI systems could understand emotions and feelings but are under development.
Self-Aware: these AI systems are designed to possess consciousness, learn independently, and form their own belief systems.
Artificial Intelligence Applications
The current developments of AI are applicable in different ways:
Self-driving cars: examples: Waymo, Tesla Autopilot, Cruise, Uber ATG, Argo AI etc.
Generative AI tools: such as ChatGPT, Gemini, Microsoft Copilot, Meta AI etc
Smart Homes
Internet of Things (IoT)
Agriculture
Finance
Education
Anomaly detection: AI systems can be used to detect an unexpected behavior
Health
Customer Service | Personal AI assistant
Research
Wearables
Google Maps
Games
Programming etc...
Subsets | Fields | Technologies in Artificial Intelligence
Machine learning: is a subset of artificial intelligence, that can adapt to many types of inputs, including large sets of past data, made-up data, or input from various sources.
Deep Learning: is an aspect of machine learning that employs complex neural networks with multiple layers. It's used in various applications like recognizing images and controlling voice commands in devices, and it excels at learning from huge datasets.
Computer Vision: enable the analysis, interpretation and extraction of information from images and videos. It can be used for facial recognition, optical character recognition, object detection, autonomous vehicles (self-driving cars).
Speech Recognition: AI systems capable of comprehending human speech, recognizing words, and discerning semantic intent.
Robotics: Robots powered by AI systems can do complicated jobs in agriculture, logistics, healthcare, manufacturing, and exploration. They can work alongside humans, learn from their experiences, and adapt to new environments.
Natural Language Processing (NLP): AI systems that can identify, analyze and understand human language. Utilizes deep learning algorithms to analyze, comprehend, and extract significance from textual information.
Generative AI: the branch of AI systems that can generate text, images, audio and videos from text instructions (prompts).
Expert Systems: the branch of AI systems that solve problems and answer questions within a specific area of expertise using rule-based methods.
Advantages of AI
Speed in execution of task regardless of their level of complexity.
Perform risky operations effectively and efficiently.
Reduce human mistakes (errors).
AI systems can handle repetitive tasks without getting fatigue.
AI systems have created more jobs such as Machine learning engineer, Prompt engineer, Data engineer, Robotics Engineer, Software engineer etc..
Limitations of AI
Development and Implementation is resource intensive and costly.
Make some jobs redundant.
Bias and discrimination can arise in AI models when they are trained on data reflecting biased human decisions, resulting in outputs that exhibit bias or discrimination against specific demographic groups.
Could violate laws or regulations across several regions.
Conclusion
The primary goal of AI development is to create self-learning systems capable of interpreting data and addressing unique challenges in ways that mirror human cognitive abilities. Artificial intelligence (AI) systems and tools significantly simplify human tasks by leveraging advanced technologies.
AI covers a wide range of technologies like machine learning, deep learning, and natural language processing. Despite the numerous potential of AI systems to boost productivity and innovation across different areas, its implementation is subject to various regulations and considerations across several regions.
Building AI systems entails extensive research, the formulation of mathematical models, and the development of sophisticated language processing capabilities. Fortunately, individuals who lack the resources to construct independent AI systems can still benefit from companies that provide AI services through APIs and other means to perform their tasks effectively and efficiently.
The advancement of computing technology once felt like the pinnacle of innovation and creativity. However, in today's world, the remarkable capabilities of smartphones have nearly matched those of traditional computers. So, in my opinion AI tools will not replace humans.
References | Resources for Further Research
Artificial Intelligence (AI) by Technology Expert Margaret Rouse
Artificial Intelligence. What Is Artificial Intelligence (AI)? Written by Ellen Glover