With the rise of tools like ChatGPT, we’re sure you’ve heard of artificial intelligence (AI). But just in case you aren’t sure exactly what it is, we’ll break it down: it’s the process of using computers for actions that normally require human intelligence. But that’s not all. It also can do more than what humans can, such as processing immense amounts of data. There’s a lot of room for improvement with AI, but the goal is for it to be able to recognize patterns, make decisions, and judge like humans can.
The history of artificial intelligence
You might think of artificial intelligence as a new trend and something that’s only come up in recent years, but that’s not true: intelligent machines were introduced as early as the 1950s. But advancement was slow due to the fact that computers were extremely expensive and not widespread; however, in the late 1950s, scientists began to dive into the possibility of artificial intelligence. It took time as people familiarized themselves with a machine with such incredible capabilities, but AI as we know it today slowly came to be.
Research slowed in the ‘70s due to a lack of funding, but the 1980s bore witness to new funds and newfound interest in the topic and deep learning came to be, impersonating for the first time the decision making process of a human. Despite this great advancement, funding decreased yet again but the possibilities of AI had grabbed the attention of the next generation of engineers and scientists: by the early 2000s, many of the goals of AI were met.
Now that we know what artificial intelligence can achieve, entire organizations are dedicated to capturing its abilities and facilitating as many of our tasks as possible - even driving! It’s constantly changing and evolving, but we can sort artificial intelligence into four different types: reactive, limited memory, theory of mind, and self-awareness. The future of AI is so wide and full of possibilities; it’s impossible to predict what’s coming next.
Artificial Intelligence: a Breakdown
Reactive artificial intelligence is the most basic and fundamental of all; it’s designed to provide a predictable answer to the input it receives and respond the same way every time; it cannot learn from past answers. This was the first iteration of AI and is therefore quite basic, but is limited in terms of what it can do. However, it has been useful in the following forms:
Spam filters to keep unwanted emails out of inboxes
Online game opponents
Recommendations on websites or streaming sites
Limited Memory AI
This next and more advanced type of AI can learn from the past, building data that evolves based on the information it receives and is the most common kind of AI found today. It’s able to understand and interpret data better than reactive AI, but still experiences limitations when it comes to storing data and using it in the future. Examples of limited memory AI are:
Predictions for the next model in a sequence
Theory of Mind AI
You might think your Alexa is able to portray emotion or that your Siri can provide you with everything you need to know, but the complexity of human beliefs and emotions means that an AI that can truly understand your emotions and provide an appropriate response is still far off. Next time Google Maps sends you in the wrong direction and you swear at it, note that it doesn't respond to your emotion. Future iterations of theory of mind AI would be able to both read and respond to human emotions.
This is the ideal level of artificial intelligence that we are still looking to achieve; a self-aware AI would be able to not only have their own emotions, but also respond to and understand the emotions of others. It will be able to make connections through emotions, such as comprehending that it is annoyed because a website won’t load.
Common uses of AI in our daily lives
AI still has a long way to go to achieve the self-aware level, but that doesn’t mean it’s not already significantly impacting our daily lives. Some of the most common uses of AI are so ingrained in our day-to-day lives that we might not even realize it’s there. Let’s take a look:
Text assistance: have you ever been drafting a text or writing on Google Docs when what you wanted to say magically appears?! It might seem like magic, but it’s just AI: here, the tool can anticipate what you want to say based on past information and make suggestions. It’s not limited to just that, however; it can also correct grammar and misspellings, grade the readability of your text, and check for plagiarism.
Social media: Instagram and TikTok are just a few of the many social media platforms that offer a page that’s personalized just for you, curated to your taste based on the pages you visit, the posts you “like,” and the accounts you follow.
Face recognition and detection technology: these two technologies are somewhat similar, recognizing a face (detection) and a specific face (recognition) to provide an enhanced user experience. Whether it’s used for unlocking your iPhone or putting a funny dog filter on your face, this technology uses previously gathered facial data to make your user experience easier.
Chatbots: recently popularized with the launch of ChatGPT, the use of chatbots for a wide range of applications is one of the most common forms of AI. Companies can train chatbots to answer FAQs, take orders, or imitate customer service representatives through natural language processing.
Search options: Google, for example, uses quality control AI software to both filter for high-quality content and what the users want, based on previous searches. Like chatbots, this uses natural language processing to properly understand user queries.
Concerns with AI
Just like with anything, there are certain aspects to keep in mind when working with AI. Even as it advances and seems to be able to create human-like responses, it will never be human and any emotions or responses will be computer-generated. And tools like facial recognition need to be limited; a misuse of such technology could mean that user privacy is violated.
It’s also essential to keep in mind that an AI’s knowledge is limited to what it’s provided with and given that tech is a rather homogeneous industry at the moment boasting primarily white, well-off, and cis-het men, it may be programmed with certain biases. This could mean AI has trouble understanding social consequences of certain things or difficulty recognizing accents or different races.
The Future of AI
Artificial intelligence has grown significantly over recent years and continues to do so, absolutely transforming the way in which we function and work. But there are things to keep in mind and the complexity of the human mind and our wide range of emotions is something that will never truly be harnessed by technology. In the meantime, however, we’re eager to see new iterations and keep learning from it.