Written By: Sophia Tang
“Hey, Alexa, what time is it?” or “Hey, Google, what’s the weather for today?” or even “Hey, Siri, set a timer for 10 minutes” are phrases we’ve all become so accustomed to in today’s society. These voice assistants are all examples of artificial intelligence (AI), which is becoming much more prevalent in our lives. For instance, taking a selfie on social media with a filter or unlocking your phone with facial recognition are both examples of how you are using AI every day. Many of us don’t even realize how widespread the influence of AI is in modern society, as it is becoming increasingly important in this digital golden age.
Artificial intelligence is an extremely broad field, usually defined as a section of computer science that allows computers to solve problems by making predictions and decisions. AI emulates human capabilities (hence its name of “artificial intelligence”), as it can provide insights or automatically trigger procedures after being presented with new data. Diagnosing diseases protecting endangered species in the wild, and providing support to locations affected by natural disasters are just a few tasks AI can be used for. In broad terms, AI is the overarching ability of computers to perform tasks by imitating human intelligence and thinking, not merely using preset algorithms. Because AI is “artificial intelligence,” it attempts to make the best decisions by what best possible actions that humans would take in the situation. Because of the complexity of the human brain, and, thus, AI, many techniques must be implemented before these technologies, such as smart devices and voice assistants, can be deployed. Tools such as machine learning, deep learning, neural networks, computer vision, and natural language processing are often used, depending on the goal of the AI machine. In this article, I’ll focus on machine learning!
Machine learning is the process of training a system to identify patterns and make decisions through large amounts of data. The machine learning algorithm is the method in which the system evaluates the data and carries out tasks, with the goal often being to predict output values from given data. To achieve the best performance and accuracy in the model, programmers must alter and fine-tune the algorithm’s parameters.
There are four broad categories of algorithms, but I’ll focus on the category of supervised learning, where the system is trained with data with inputs and outputs. After being trained on data containing inputs as well as outputs, the goal of the algorithm is to discover the method to arrive at the most accurate output given only the inputs. Supervised learning can be further split into classification, regression, and forecasting.
Classification algorithms group data into subgroups; for instance, one popular implementation is when Gmail filters emails into “spam” and “non-spam.” It is important to note that sometimes Gmail will mark an important email as spam or vice versa due to imperfections in the algorithm, insufficient data, or many other factors, meaning machine learning isn’t perfect, but it can be improved.
Regression algorithms model the correlation and relationship between variables to predict a continuous value instead of a category. One common example is estimating real estate value regarding location, lot size, number of bedrooms, etc. The real estate price cannot be categorized because it is a continuous value.
Forecasting algorithms focus on predicting the future, given past and present data, such as weather reports, which are a common tool in many of our lives.
So why are machine learning and artificial intelligence so important? Most successful companies collect data and then utilize machine learning models to provide insightful information on how to optimize their products, market flow, etc.
I’m sure all of us are familiar with YouTube, TikTok, and other similar platforms. How do they recommend such accurate videos, knowing we will click on them and thus create more traffic for their websites by spending more time on them? They use machine learning! These applications track the videos you watch, record the amount of time you watched them, note down if you liked or disliked them, and many other factors. Using this data, they can predict which videos you will be more inclined to click on, and they put these on your “For You” page.
But machine learning applications are not only restricted to the entertainment industry. AI is increasingly important for security, especially in banking. Machine learning is used to detect and prevent fraud and correctly authenticate users and documents such as checks. Machine learning also often aids doctors’ decisions by analyzing patient health records and assisting in medical imaging. Another prospecting aspect for the future is using machine learning to help diagnose certain diseases, such as cancer.
With machine learning and AI becoming so popular and widespread, they impact our lives in so many ways. And although we still have a long journey ahead with issues like ethics and more advanced development of this technology, AI is becoming increasingly noticeable in almost every area of our life and has much potential for the future to solve global issues.
Comments