In The Wizard of Oz, when Dorothy and her companions finally meet the Wizard, they are initially confronted by a formidable presence: a giant head, a ball of fire, and other intimidating illusions. However, Dorothy's dog, Toto, pulls back a curtain to reveal that these spectacles are the work of an ordinary man operating machinery and speaking into a microphone. This man, who had been presenting himself as the powerful Wizard, is a regular person using tricks to maintain an illusion of grandeur.
Even though Artificial Intelligence is not merely an illusion, I’ll pull back the curtain on how AI works and some of the differences between machine and generative learning. Because AI is already being used to tackle complex challenges and improve efficiency in a variety of industries, I feel it’s very valuable to know the basics of how it works. Today, most of us think nothing of using calculators instead of an abacus or pencil and paper to calculate figures. The simple calculator has become a normal part of our routine. In the same way, many aspects of AI are already integrated in many parts of our daily lives. First, I’m going to talk about input, then distinguish between machine learning and generative learning.

Input for AI Platforms
When we memorize a spelling list, we reinforce each word in our memory through repetition and practice. Most of those words are stored in our brain’s memory systems to be retrieved when writing or speaking. Very similarly, all AI models are trained on large databases that are inputted then stored. Then, that information is retrieved for content generation. Both processes involve recognizing and internalizing patterns; humans learn the sequence of letters in words and AI learns patterns in data. Repetition enhances proficiency in both situations of a spelling list and AI output.
Input is crucial in the growth and effectiveness of AI, as it directly influences the learning process and outcomes. The quality and quantity of data fed into AI models are foundational to their development. For example, machine learning models learn from vast datasets, and the more diverse, accurate, and relevant the input data is, the better the AI can perform. Over the years, significant amounts of data have been used to train AI systems, with billions of data points processed from various sources like social media, healthcare records, customer behavior, and more.
The statistic that over 2.5 quintillion bytes of data are created daily is widely cited across various sources, including Forbes, providing a constant flow of input for AI models. As this data continues to grow, so too does the potential of AI to improve decision-making, automate tasks, and even predict future trends. It's important to note that data generation rates are continually increasing, with estimates suggesting that by 2025, the amount of data generated each day will reach 463 exabytes globally. 463 exabytes is the number 463 with nine zeros. That’s a lot of data to be stored and ready for access. (Haslam College of Business)
Types of Learning for AI
We focus here on machine learning and generative learning, though AI continues to evolve with new terms and learning methods. However, most advancements stem from these two core principles.
Machine learning is a branch of artificial intelligence (AI) that enables computers to learn and improve from experience without being explicitly programmed. It uses algorithms, which are a set of predetermined rules, to analyze large amounts of data, identify patterns, and make predictions or decisions based on that data. For example, machine learning powers recommendations on streaming platforms, fraud detection in banking, and voice recognition in virtual assistants. All of those recommendations are based on the step-by-step instructions taken from a large amount of data. The key to machine learning is that it continually adapts and improves as it processes more data, making it a powerful tool for solving complex problems and automating tasks in diverse industries.
Deep learning is a specialized subset of machine learning modeled after the human brain that excels at processing large amounts of complex data, such as images and speech and is the driving force behind facial recognition and language translation.
If you are in the field of finance or medicine, machine learning is extremely useful in performing calculations and determining patterns. Those calculations that in the past may take weeks, can now be done in seconds. The information gathered doesn’t make decisions for us but provides the tools for us to make better decisions.
Generative learning is a subset of machine learning focused on creating new data that resembles the original data it was trained on. Unlike traditional machine learning models that focus on predictions or classifications based on data, generative learning creates something entirely new by learning patterns and structures from existing datasets. As an example, I have created illustrations for articles with Canva’s version of generative learning. Some illustrations are usable and some are not. Much of the outcome depends on the type of prompt I give for the illustration. The more specific, the better the outcome. Generative learning has immense potential for creativity, design, and innovation across industries, enabling businesses to automate creative processes or develop unique solutions tailored to their needs.
One of the huge issues with generative learning is copyright protection as the generative product is based on existing patterns and structures. This will be an ongoing conversation with the analyzation of how close new data is to previously protected data, especially relevant for those with intellectual property. This should not hold us back from using this type of AI, but it should make us aware and vigilant with its use.
Enhanced Efficiency and Productivity
As mentioned, AI can automate repetitive tasks, streamline workflows, and provide data-driven insights, allowing businesses to operate more efficiently and focus on strategic activities. Below are some examples.
Healthcare: AI is especially valuable in early disease identification. This area is growing with the access of more and more data. Soon there will be tests for even some of the issues that stay hidden until it’s too late for treatment, such as cervical cancer. The American Cancer Society estimates that in the United States in 2025, 13,360 new cases of invasive cervical cancer will be diagnosed and about 4,320 deaths. * It’s known as the fourth most common cancer in women globally.
Can you imagine the emotional impact of walking in to your doctor and obtaining an early diagnosis that will save your life? I believe this type of AI expansion in the medical field will help to save lives and bring medical costs down with a quicker, accurate diagnosis and resulting treatment.
Healthcare currently uses A.I. to analyze patient data and medical literature to provide doctors with evidence-based treatment options, especially for cancer. This fact highlights the importance of quality data input. Just like performing a search on Google, it’s important to know and trust the source of information. Data that used to take months and even years to analyze is now performed immediately, allowing researchers to build upon that data. This area will continue to grow.
Fraud Detection: In finance, A.I. has become a pivotal tool in detecting fraudulent activities. By analyzing vast amounts of data, when deviations of normal patterns occur, the system flags them. This is why we get those notices to approve a particular purchase if it falls outside our normal patterns of purchase. Maybe you, like I, have had this occur when traveling and the vendor isn’t easily identified.
It can be particularly annoying when a transaction doesn’t go through promptly while traveling, but ultimately, it will save both the card holder and the company time and resources. It’s frustrating when my card is cancelled as it takes time to go back through every automated payment to update my card. However, I know the protections are in my best interest and will save both me and the credit card company valuable time and money.
As consumers of AI tools, the challenge lies in staying open to learning and strategically adopting tools that align with specific business needs. The key is to focus on solutions that provide the greatest value relative to the time and energy required for implementation. I carefully evaluate the time it will take to use certain tools as the learning curve and implementation may not be worth the effort. Additionally, I know that by next week or month there will be another tool that may serve my needs better. But that also doesn’t mean that I hesitate too long. Most choices I have personally made have ultimately worked well.
By making thoughtful choices, businesses can capitalize on AI’s capabilities while ensuring it serves their unique goals effectively.
Application
To start this process, pursue the following three steps that will open up your thoughts:
STEP ONE: Read and research about the different types of Artificial Intelligence. The main types discussed here are the basics. There is the input of data, then analyzation and extraction of that data applied to a prompt.
STEP TWO: Identify the current skillset and work you do that could benefit from a certain type of Artificial Intelligence. Ideas are for the medical, organizational, financial or creative fields. The possibilities are endless.
STEP THREE: If you are already using some AI tools, identify others that will be the most beneficial. Consider this carefully with tools that will ultimately save time and resources.
Machine learning models learn from vast datasets, and the more diverse, accurate, and relevant the input data is, the better the AI can perform. Deborah Johnson
Additional Sources
Additional studies on AI and Cancer: Cancer Studies
Artificial Intelligence in Healthcare – This article is more than 6 years old, but holds some good basic information.
The Age of AI: And Our Human Future by Kissinger, Schmidt and Huttenlocher
Inside AI: Over 150 billion purchases per year use this author's AI by Akli Adjaoute
1,560 words