The AI Handbook: A Casual Approach to Understanding Artificial Intelligence
Your Beginner’s Guide to AI.
Artificial intelligence (AI) has become a familiar presence in our lives, quietly working behind the scenes. While concerns and discussions about AI’s impact and ethical considerations are present, our primary goal is to provide beginners with a comprehensive understanding. Before we explore the latest developments and possible concerns, let’s start with the basics. In this guide, we’ll unravel what AI is, its various types, and the exciting possibilities it brings, making this an ideal starting point for those new to the world of AI.
To kick off our discussion, let’s delve into the very first question:
What exactly is AI?
Artificial intelligence, often abbreviated as AI, involves computer tasks capable of performing and evolving independently without the need for human interaction or assistance. These functions include learning, problem-solving, reasoning, and recognition.
Let’s explore this further with the types of AI and some practical examples.
Reactive Machines
Reactive machines are the most basic type of artificial intelligence. They are developed with programmed instructions and cannot perform tasks beyond their limited capabilities. Instead of remembering or storing data, they can only perform tasks pre-programmed into them.
The best examples of reactive machines are:
- Chess-Playing Programs: They rely on pre-programmed rules and strategies to make their next move. However, in 1989, Chess Master Gary Kasparov defeated IBM’s Deep Thought, a computer designed to play chess. On the other hand, in 1997, IBM’s Deep Blue defeated the same chess master, marking a pivotal moment in the history of human-AI interactions.
- Calculators: They can perform basic operations following pre-programmed rules but can’t handle complex problems or new mathematical operations.
- Thermostats: These machines respond to the current location’s temperature and maintain the specified range. They have been helpful, especially with the number of thermostats in homes.
- Traffic Lights: These machines operate based on pre-defined sensors and patterns, classifying them as reactive machines. Similar to the limitations of reactive machines, traffic lights do not adapt to constantly changing traffic conditions in real time.
Limited Memory Machines
Limited Memory Machines possess a restricted grasp of past events. In contrast with reactive machines, this type of AI can engage more dynamically with its surroundings. However, due to their limited memory, their recollection of past events remains within a narrow timeframe. Some examples are:
- Virtual Personal Assistants: Some of the most popular virtual personal assistants are Google, Siri, and Alexa. They can answer questions, set reminders, and perform tasks based on human input. Despite their limited memory, they can recall previous commands to provide context for future interactions.
- Recommendation Systems (Netflix, Shopee, and Advertisements): Entertainment, e-commerce, and social media platforms use data from a user’s previous selections and preferences to provide recommendations. This principle also extends to online advertisements, helping advertisers tailor their delivery based on customer interests and behavior, including the websites they visit and the products they view.
- Predictive Text and Autocorrect: As you type on your smartphone, it saves the most common words or phrases to enhance user convenience. This type of AI relies on user input but does not learn or adapt beyond preset settings.
- Basic Chatbots for Support: Many website chatbots feature predefined sets of questions, including frequently asked questions (FAQs), tailored to offer sophisticated and personalized customer support. These chatbots can remember past conversations and provide context-aware responses based on your previous interactions. Additionally, they serve multiple functions, often acting as appointment schedulers, weather or news bots, and language translation tools.
Narrow AI
Narrow AI (or Weak AI) and Limited Memory Machines share some similarities. They both operate within specific domains and have limitations compared to general AI. While their contexts overlap, they are not exactly the same.
This type of AI may not have limitations related to data storage; instead, they are typically defined by the tasks and functionalities they are designed to perform, such as speech recognition, image identification, and translation.
Some of the examples also belong to the category of limited memory machines, but they are more advanced. Here, I will provide more examples typically categorized as narrow AI:
- Image and Speech Recognition: AI systems like these are typically used in facial recognition, content tagging, and speech-to-text conversion. This kind of AI is being used in airports, building entry barriers, and even detecting anomalies in medical images like X-rays or even in analyzing medical data.
- Autonomous Vehicles: With the increasing number of new car models in the industry, technology is, of course, present, and there are also increased mind-blowing features. However, they are usually specialized for driving tasks, just as airplanes are on autopilot.
- Gaming: The opponents you typically encounter in video games are, indeed, AI. Types of games like strategy games, role-playing games (RPGs), and simulation games also incorporate AI to make the gameplay more engaging and challenging for the player.
- Art, Music, Image, Video, and Literature Generation: While this has become a pressing issue in recent times, AI algorithms can assist users in analyzing and mimicking art styles, compositions, coherent literary texts, and even video editing techniques from various references, enabling them to create new art pieces and edit multimedia content. This capability extends to photo and video editing applications like Photoshop and video editing software.
- Email Filtering: Since their main purpose is to categorize and separate spam emails from the important ones, they fall under the category of Narrow AI. They have training and specifications and are task-specific despite their limited scope.
Theory of Mind Machines
These AI systems represent an early and evolving stage of artificial intelligence. Theory of Mind machines aims to understand the emotions, intentions, and beliefs of entities in their environment to improve human and AI interactions. Despite these ambitions, achieving a fully developed Theory of Mind machines remains challenging in AI research.
Let’s take a closer look at some practical examples of Theory of Mind AI machines that are pushing the boundaries of human-AI interaction:
- Social Robots: Are you familiar with advanced social robots like Sophia and Pepper? They possess the ability to recognize and engage in emotionally intelligent interactions with humans, thanks to the sensors and algorithms embedded in their design. Within the category of social robots, this technology extends to humanoid companions, which are utilized for elderly care and companionship, assisting in empathetic interactions.
- Emotionally Aware Virtual Assistants: In this technology-dominated generation, Amazon’s Alexa and Apple’s Siri have improved over the years, becoming more context-aware and capable of providing appropriate responses based on tone and conversation. They are also widely used on devices such as smartphones and Alexa-enabled devices.
Self-Aware Machines
Now, this concept remains a distant reality. Self-aware machines represent the most advanced type of AI in theory. They would possess a profound understanding of the world, people, and themselves — territory that science fiction novels often explore. Here are some characteristics that a self-aware machine could have:
- Consciousness
- Emotional Intelligence
- Creativity and Original Thought
- Meta Cognition
How does AI work exactly?
We have various types and examples explored above, and it becomes clear that AI is a pervasive presence in our lives. However, as we delve into these examples, it’s natural to wonder about the inner workings of AI. How did it come into existence? What were the origins and motivations behind its development?
Machine Learning:
This section of AI primarily focuses on developing algorithms and statistical models that make machines appear alive. In reality, it consists of models enabling them to learn and make predictions without requiring overly complicated programming. Machine learning helps healthcare, finance, natural language processing, recommendation systems, and autonomous vehicles find their applications.
Deep Learning:
This process is where artificial and deep neural networks excel in modeling complex patterns. Deep learning, a subfield of machine learning, has led to advancements in computer vision, natural language understanding, speech recognition, and autonomous robotics.
Tips and Bits of Advice to get you started
You want to dive into the world of artificial intelligence. Good for you! Here are five pieces of advice you can follow:
- Learn the Basics: Of course, the journey of a thousand miles begins with the first step. Along with this article as your guide in artificial intelligence, it’s essential to understand its fundamental concepts.
- Take Online Courses: Gain practical skills. It’s vital to apply what you have learned in practice. That’s what it takes to gain further understanding.
- Work on Projects: Application! You’re now equipped with the knowledge you need in AI. Apply it in hands-on projects, whether for yourself or others, to build experience. Teaching is also a great way to remember everything you have learned. After all, applied knowledge is wisdom.
- Dabble in Programming: If you want to take things further in AI and become one of the developers working behind the scenes, get into programming. The common languages used in AI are Python, C++, and Java.
- Stay Updated: Information and data are constantly changing, and something new happens every day. It’s best to keep in touch with it as well. Engage with communities, research, and challenges.
Conclusion
As we wrap up this guide, we’ve covered the basics and shown examples of how AI works in many ways. AI is becoming a big part of our lives, so it’s necessary to understand its basics and ethics. Whether you’re using AI for any reason, there are many opportunities for those curious to explore its possibilities and be part of its growth.
Additional Resources:
- Ever wondered how AI would perform at human jobs? Watch this video by WIRED on YouTube as AI Tries 20 Jobs.
- Diving deeper? Here are FREE AI Courses in Udemy.
- Get certified with this Artificial Intelligence Program from Stanford Online.
Learn more about SparkLearn EdTech from the following links:
E-learning Platform: lrn.ac
Twitter: lrn.ac/twitter
Facebook: lrn.ac/facebook
Instagram: lrn.ac/instagram
LinkedIn: lrn.ac/linkedin
YouTube: lrn.ac/youtube
