What Is Artificial Intelligence?
Artificial intelligence (AI) has been mentioned all over the news lately, but what exactly is it? I asked ChatGPT to explain AI in simple terms, and this is the response it generated:
“AI, or Artificial Intelligence, is a field of computer science that focuses on creating smart machines capable of performing tasks that typically require human intelligence.
Think of AI as a computer program or system that can learn, reason, and make decisions based on data and patterns it has been trained on. It’s like giving a computer the ability to think and make sense of information, just like we do.
AI uses various techniques, such as machine learning, to train computers to recognize patterns in large amounts of data. By analyzing this data, AI systems can make predictions, solve problems, and even perform tasks like image recognition, language translation, or playing games.
One common type of AI is called “narrow” or “weak” AI. This refers to AI that is designed to perform specific tasks really well, but it doesn’t possess general intelligence like humans do. For example, a voice assistant like Siri or Alexa is a type of narrow AI.
On the other hand, “strong” AI, also known as “general” AI, is the type of AI that has the ability to understand, learn, and reason across a wide range of tasks, similar to human intelligence. However, we haven’t achieved strong AI yet, and current AI systems are still far from having human-like understanding and consciousness.
Overall, AI is about creating smart computer systems that can analyze data, learn from it, and make decisions or perform tasks that would typically require human intelligence. It has the potential to revolutionize various industries and improve our lives in many ways.”
Of course, ChatGPT itself is a popular example of an AI program called a “chatbot.” The answer it generated is coherent, but does it really tell you exactly what AI is? Let’s clarify a few points.
Artificial Intelligence: Theory vs. Application
The first paragraph ChatGPT provided mentions “smart machines capable of performing tasks that typically require human intelligence,” which might conjure images of highly advanced robots like those in I, Robot or Ex Machina. It clarifies this point, however, in the first sentence of the second paragraph by describing AI as a computer program. Let’s start there.
When AI is mentioned in the news or in conversation, what people are typically talking about is AI computer programs. While the field of AI is more generally the science and engineering of making intelligent machines, the applications of AI are currently manifested in a variety of AI computer programs.
An AI program is written by human programmers using a variety of high-level computer programming languages (such as Java, Python, or C++). These languages provide a more human-readable and expressive syntax, which is then interpreted into binary code (a series of ones and zeros) for the computer to execute. Essentially, programmers write a program in high-level code to design and implement instructions and parameters about how the program will function and operate.
Does Artificial Intelligence Really “Learn?”
In the third paragraph, ChatGPT tells us that “AI uses various techniques, such as machine learning, to train computers to recognize patterns in large amounts of data.” Before we define “machine learning,” let’s go over another important definition.
You may have heard of the word “algorithm” associated with AI or other computer programs. Popularly, social media algorithms are used to sort posts based on patterns and preferences of users rather than in chronological order as they are posted.
In the field of AI, an algorithm can be synonymous with an AI program. An algorithm is a set of instructions or rules that define a sequence of steps to perform a task, also known as the AI program in its form as written computer code. There are different kinds of AI algorithms that vary in complexity and purpose, but the basis of AI programs revolves around the use of algorithms.
Machine Learning Algorithms
A specific type of algorithm is applied in “machine learning,” which enables computers to learn and improve from data without being explicitly programmed. Instead of programming all of the rules explicitly, machine learning algorithms learn patterns and make predictions based on data.
Machine learning starts with data, usually a lot of data, which can be pictures or text. The algorithm is trained on this data by being shown the data with instructions on what to recognize. During training, the algorithm adjusts its internal parameters based on the patterns it discovers in the data, refining how the algorithm analyzes the data. Once the algorithm has been trained, when it is presented new never-been seen before data, it makes predictions or decisions about this data.
Weak AI vs. Strong AI and the Advancement of Artificial Intelligence
Finally, ChatGPT describes “weak” or “narrow” AI programs and “strong” or “general” AI. Although ChatGPT suggests that “a voice assistant like Siri or Alexa is a type of narrow AI,” so too is ChatGPT an example of a weak or narrow AI. In its own words,
“ChatGPT can be considered an example of weak AI. Weak AI, also known as narrow AI, refers to AI systems that are designed to perform specific tasks or functions really well but do not possess general intelligence or consciousness like humans. ChatGPT is a language model that excels at generating human-like responses based on the input it receives.”
AI programs are highly advancing, but “strong” or “general” AI programs with human-like consciousness have not yet been developed. Overall, though, ChatGPT was able to generate a coherent and largely accurate response. With the addition of some human explanation, this article has hopefully helped you understand how AI programs work.
This article was written by Avisen Law Clerk Javier Romay, with ChatGPT, to demonstrate how AI programs perform and can be explained, under the direction of Avisen Founder Kim Lowe.
